Issuu

iafor

journal of education

technology in education Volume 14 – Issue 2 – 2026 Editor: Michael P. Menchaca

ISSN: 2187-0594

iafor IAFOR Journal of Education: Language Learning in Education Volume 14 – Issue 2 – 2026

IAFOR Publications The International Academic Forum

IAFOR Journal of Education: Technology in Education Editor Michael P. Menchaca University of Hawai‘i at Mānoa Associate Editor Daniel L. Hoffman University of Hawai‘i at Mānoa Associate Editor Devayani Tirthali Research Associate, India Published by The International Academic Forum (IAFOR), Japan IAFOR Publications, Sakae 1-16-26-201, Naka-ward, Aichi, Japan 460-0008 Publications & Communications Coordinator: Mark Kenneth Camiling Publications Manager: Nick Potts IAFOR Journal of Education: Technology in Education Volume 14– Issue 2 – 2026 Executive Editor Joseph Haldane The International Academic Forum, Japan Published August, 2026 IAFOR Publications © Copyright 2026 ISSN: 2187-0594 ije.iafor.org

IAFOR Journal of Education: Technology in Education Volume 14 – Issue 2 – 2026 Edited by Michael P. Menchaca Associate Editor: Daniel L. Hoffman Associate Editor: Devayani Tirthali

Table of Contents

From the Editors Michael P. Menchaca, Editor Daniel L. Hoffman, Associate Editor Devayani Tirthali, Associate Editor

Contributors

Exploring an Educator’s Experience in Higher Education During Generative AI Transformation Karen K. Fujii Natalie Perez

Exploring Hidden Structure and Improvised Practices Shaping Student Collaboration and Agency in Online Learning Farha Alia Mokhtar Sally Barnes

Prompting Intentionality and Authorship Preservation in AI-Assisted EFL Academic Writing: A Process-Tracing Inquiry Rym Ladjal Hayat Messekher

Applying Gamification to Undergraduate Accounting Education: A Mixed-Methods Study Tialei Scanlan Spencer Scanlan

Relationship Between School-Related and Private Digital Media Use and Self-Learning Competences Among German Vocational Students Maxi Eileen Brausch-Böger Manuel Förster

117

AI Literacy and Convergence Competencies Among Pre-Service Teachers Juyoung Lee

147

Gamifying Sepedi Riddles for Developing Critical Thinking in Grade 7 Learners Nkame Emmanuel Ngobeni Ablonia Dihloriso Maledu

173

Four-Domain Framework for Evaluating Education Management Information System Effectiveness in Bhutan Gembo Tshering Saroj Thapa Tenzin Choden Lekphell Chimi Dema

195

Reviewers

226

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

From the Editors Welcome to the 2026 IAFOR Journal of Education: Technology in Education issue. We, the editors, note that the post-pandemic years have brought tremendous, almost frenetic, adoption of varied technologies as emergency remote measures transform into a “new normal”. Researchers and educators no longer merely adapt out of necessity; they actively adopt and analyze, evidenced by the rise of increasingly complex learning spaces. This issue reflects that transformation by showcasing a collection of eight international articles representing an array of rigorous methodologies, sound theoretic foundations, and innovative technologies. Reinforcing IAFOR’s commitment to international, intercultural, and interdisciplinary exchange, the manuscripts represent diverse studies from four continents and eight countries including Denmark, Malaysia, Algeria, Germany, South Korea, South Africa, Bhutan, and the United States. Methodologies include qualitative, quantitative, mixed-method, and design-based inquiries and are supported by sound theoretic foundations. As editors, we receive numerous submissions exploring novel tools with interesting educational applications. Many are engaging and valuable. However, we consistently emphasize that meaningful research lies not in the mere adoption of a tool, but in its grounding within robust frameworks. The articles in this issue demonstrate exemplary practice by aligning questions, methods, findings, conclusions, and implications within frameworks such as transformative learning theory, sociocultural theory, mediated action, self-determination theory, systems evaluation, and technological pedagogical content knowledge. We encourage our readers and especially prospective authors to reflect on how these frameworks guide the presented research. The technologies represented across these articles span an intricate continuum stretching from localized collaboration tools to gamified cultural applications to cutting-edge GenAI models to macro-level national database systems. Rather than treating these tools as sole applications, the articles collectively highlight how technology is adopted and negotiated within contemporary educational spaces to support student self-regulation, preserve human agency, and enable data-driven governance. When perusing the articles, readers may note three emerging trends in the contemporary educational technology landscape: •

•

Human agency and authorship in the era of AI. As GenAI tools become more commonplace, educators and student writers are actively renegotiating what it means to think, write, and create academic work. The articles in this issue move beyond conventional debate pitting utopian adoption against dystopian resistance and instead explore cognitive effort, professional identity construction, and exact mechanisms for prompting. Gamification and culturally-situated motivation. Game elements such as badges and leaderboards continue to hold immense promise for student engagement. However, the articles indicate gamification is not a panacea. Success depends on the dynamics of competition, the perceived value of reward, and the integration of cultural awareness to foster critical thinking.

IAFOR Journal of Education: Technology in Education

•

Volume 14 – Issue 2– 2026

Data-driven education and self-regulation. Contemporary learning spaces are fundamentally altering how students regulate their own learning and how institutions monitor that learning as well as overall educational quality. From the self-learning competencies of vocational students in Europe to the national integration of learning management platforms in Asia, the articles show technology is transforming data into actionable pedagogy.

Together, the eight articles demonstrate that integrating technology is not a straightforward transfer of resources to some abstract environment, virtual or otherwise. Rather, the innovative utilization of technology requires deliberate, culturally honest, and theoretically sound pedagogical orchestration. Whether exploring the GenAI user mind (Article 1), assessing student agency (Article 3), evaluating the moral dimensions of gamified folklore (Article 7), or building robust national information systems (Article 8), the scholarship provided here serves as a powerful reminder: The central force in educational technology remains fundamentally human. We hope you find these articles as illuminating, thought-provoking, and inspiring to read as we did to review and curate. Brief overviews of the eight manuscripts featured in this issue, organized by order of publication, follow below. Happy thinking! Michael P. Menchaca, Editor; Daniel L. Hoffman and Devayani Tirthali, Associate Editors IAFOR Journal of Education: Technology in Education Email: tech.editor.joe@iafor.org Articles Article 1: Exploring an Educator’s Experience in Higher Education During Generative AI Transformation. Authors: Karen K. Fujii (Niels Brock Copenhagen Business College, Denmark) and Natalie Perez (Private Corporation, Denmark) This qualitative study employs Interpretative Phenomenological Analysis to explore the lived experiences of ‘Thomas’, a curriculum leader and instructor with 15 years of higher education experience, navigating rapid GenAI policy shifts at a private European university. Over successive semesters, his institution’s position toward GenAI shifted dramatically from viewing it as dangerous to distracting and finally to essential. Thomas’ journey revealed deep tensions navigating a collaborative but pedagogically divided teaching team where some peers rejected AI entirely while others adopted it exclusively. Ultimately, Thomas’ experience illustrated a shift from merely checking student authorship to evaluating how students critically analyze, edit, and defend AI-generated output, treating them more as editors than traditional writers. Article 2: Exploring Hidden Structure and Improvised Practices Shaping Student Collaboration and Agency in Online Learning. Authors: Farha Alia Mokhtar (Universiti Malaysia Terengganu, Malaysia) and Sally Barnes (University of Bristol, United Kingdom)

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Framed through a Vygotskian sociocultural perspective, this qualitative case study examined how final-year Malaysian undergraduates enrolled in a mandatory English language course negotiated peer collaboration in a fully online learning environment. Students collaborated within a mandated virtual learning environment along with voluntary platforms like Google Classroom, Padlet, and WhatsApp. Researchers traced how students improvised “structured social space” to complete group tasks within these environments. Findings revealed structured but non-linear workflows where students: (a) allocated tasks first-come, first-served, (b) relied on volunteerism for organization, and (c) sought non-supported, third-party applications such as Padlet for peer-checking and other tasks. Such self-organized agency carried clear constraints including uneven peer participation, issues with internet connectivity, and, most significantly, a tendency to prioritize task submission over deep content comprehension Article 3: Prompting Intentionality and Authorship Preservation in AI-Assisted EFL Academic Writing: A Process-Tracing Inquiry. Authors: Rym Ladjal and Hayat Messekher (École Normale Supérieure de Bouzaréah, Algeria) Addressing a critical gap in the empirical research on GenAI in African higher education contexts, this study explored how 16 Algerian English as a Foreign Language (EFL) doctoral researchers negotiated authorship while using LLMs for academic writing. Utilizing an innovative five-stage process-tracing protocol, the researchers assessed how “prompting intentionality”, or the degree to which writers encode reasoning, stance, and epistemology when creating prompts, affected their preservation of “intellectual DNA”. The study found that prompting intentionality rose significantly across measured stages and that rise was strongly associated with human agency preservation. The authors propose an “Agentic Prompting Loop” to describe the co-movement of prompting intentionality, epistemic authority, and authorship preservation. Article 4: Applying Gamification to Undergraduate Accounting Education: A Mixed-Methods Study. Authors: Tialei Scanlan and Spencer Scanlan (Brigham Young University–Hawaii, USA) This embedded mixed-methods study examined the effect of digital badging and leaderboards among 202 introductory undergraduate financial accounting students. Grounded in SelfDetermination Theory, quantitative findings indicated that gamified elements kept student motivation high during typical “mid-semester slumps”; however, overall academic performance and behavioral engagement showed no significant differences across groups. Interestingly, the intervention was moderated by gender: (a) male students in the badge groups exhibited significantly higher intrinsic motivation than those in the leaderboard-only control, whereas (b) female students in the leaderboard-only control exhibited higher intrinsic motivation than those in the badged. Qualitative findings added depth as students described positive outcomes like progress tracking and chunking content, as well as challenges such as amotivation stemming from digital badges feeling like “just pixels” and lacking tangible course value.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Article 5: Relationship between School-Related and Private Digital Media Use and SelfLearning Competences among German Vocational Students. Authors: Maxi Eileen BrauschBöger and Manuel Förster (Technical University of Munich, Germany) This quantitative cross-sectional study of 248 German vocational trainees investigated how different types of digital media use related to five dimensions of Self-Learning Competence: professional, methodical, personal, emotional, and social. Structural Equation Modeling revealed that school-related media use (e.g., lesson preparation) was positively and significantly associated with methodical and personal competence, as well as emotional and social in the broader model. Conversely, private media use (e.g., social networks) was negatively and significantly associated with methodical competence. Surprisingly, no forms of media use were associated with professional competence, underscoring that while digital media supports the self-regulatory and motivational processes of learning, it should not replace the hands-on, practical expertise central to vocational education. Article 6: AI Literacy and Convergence Competencies Among Pre-service Teachers in NonSTEM Disciplines. Author: Juyoung Lee (Seoul National University of Education, Republic of Korea) This study addressed a critical gap in teacher education literature by examining levels of AI literacy and AI convergence competencies among 112 Korean pre-service teachers majoring in non-STEM disciplines. Utilizing an AI Literacy Diagnostic Tool and an AI Convergence Competencies Scale, the study found that while overall scores rose above the scale midpoint, critical technical gaps remained. Additionally, data literacy and basic AI knowledge sat at the midpoint, while basic programming scored significantly lower. Also, meaningful demographic differences emerged. For example, female candidates scored significantly higher in basic knowledge, social impact, technology utilization, problem-solving, and ethics while humanities/social science majors outperformed arts/physical education majors in social impact and ethics domains. Overall, the study found value-oriented dimensions (e.g., ethics and openness) did not correlate with technical competency, indicating these competencies should be assessed and reported as discrete profiles rather than aggregated composites. Article 7: Gamifying Sepedi Riddles for Developing Critical Thinking in Grade 7 Learners. Authors: Nkame Emmanuel Ngobeni and Ablonia Dihloriso Maledu (University of Limpopo, South Africa) Grounded in Vygotskian sociocultural theory, Anderson and Krathwohl's revised taxonomy, and the TPACK framework, this design-based research study reported on the gamification of ten Sepedi riddles (dithai). Run across three iterations with four Sepedi Home Language teachers and twenty Grade 7 learners in rural South Africa, the study deployed a mobilefriendly progressive web application via Netlify to bypass typical local connectivity challenges. The browser-based game presented riddles across ascending cognitive levels. Findings demonstrated that the game: (a) successfully sustained learner engagement with higher-order

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

reasoning, (b) stimulated culturally grounded reasoning using indigenous ethical frameworks, and (c) strengthened the home-school connection through multigenerational play in the home. Article 8: Four-Domain Framework for Evaluating Education Management Information System Effectiveness in Bhutan. Authors: Gembo Tshering (Royal University of Bhutan, Bhutan), Saroj Thapa (Druk Gyalpo’s Institute, Bhutan), Tenzin Choden Lekphell, and Chimi Dema (Royal University of Bhutan, Bhutan) This cross-sectional evaluative study assessed the effectiveness of Bhutan’s national Education Management Information System alongside its homegrown, qualitative “Motherboard” assessment platform using a systems framework. Drawing on survey data from 281 schools, 20 district officers, and national policymakers, the study identified significant discrepancies between system design and operational reality. While Data Quality and Utilization in Decision Making were rated at the established level, Enabling Environment and overall System Soundness remained only at the emerging level due to infrastructural constraints, legal framework bottlenecks, and limited professional development. The paper highlights the value of combining quantitative administrative metrics with qualitative, student-centered learning analytics and outlines Bhutan’s ongoing national integration project designed to resolve data silos.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Authors List Article 1: Exploring an Educator’s Experience in Higher Education During Generative AI Transformation Karen Fujii Niels Brock Copenhagen Business College, Denmark Email: karenkf@hawaii.edu Natalie Perez Independent Scholar, Denmark Article 2: Exploring Hidden Structure and Improvised Practices Shaping Student Collaboration and Agency in Online Learning Farha Alia Mokhtar Universiti Malaysia Terengganu, Malaysia Email: alia.mokhtar@umt.edu.my Sally Barnes University of Bristol, United Kingdom Article 3: Prompting Intentionality and Authorship Preservation in AI-Assisted EFL Academic Writing: A Process-Tracing Inquiry Rym Ladjal Laboratoire de Linguistique et Sociodidactique du Plurilinguisme LISODIP. École Normale Supérieure de Bouzaréah- Chikh Moubarek BenMohammed Ibrahimi Elmili Eldjazairi, ENSB. Algiers, Algeria Email: rym.ladjal@ensb.dz Hayat Messekher École Normale Supérieure de Bouzaréah - Chikh Moubarek Ben Mohammed Ibrahimi Elmili Eldjazairi, Algeria

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Article 4: Applying Gamification to Undergraduate Accounting Education: A Mixed-Methods Study Tialei Scanlan Brigham Young University–Hawaii, United States Email: tialei.scanlan@byuh.edu Spencer Scanlan Brigham Young University–Hawaii, United States Article 5: Relationship Between School-Related and Private Digital Media Use and Self-Learning Competencies Among German Vocational Students Maxi Eileen Brausch-Böger Technical University of Munich, Germany Email: maxi.brausch@tum.de Manuel Förster Technical University of Munich, Germany Article 6: AI Literacy and Convergence Competencies Among Pre-Service Teachers in Non-STEM Disciplines Juyoung Lee Seoul National University of Education, Republic of Korea Email: leejysam@daum.net Article 7: Gamifying Sepedi Riddles for Developing Critical Thinking in Grade 7 Learners Nkame Emmanuel Ngobeni University of Limpopo, South Africa Email: nkamengobeni@gmail.com Ablonia Dihloriso Maledu University of Limpopo, South Africa

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Article 8: Four-Domain Framework for Evaluating Education Management Information System Effectiveness in Bhutan Gembo Tshering Paro College of Education, Royal University of Bhutan, Bhutan Email: gembotshering.pce@rub.edu.bt Saroj Thapa Royal Academy, Druk Gyalpo’s Institute, Bhutan Tenzin Choden Lekphell Paro College of Education, Royal University of Bhutan, Bhutan Chimi Dema Paro College of Education, Royal University of Bhutan, Bhutan

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Exploring an Educator’s Experience in Higher Education During Generative AI Transformation Karen K. Fujii Niels Brock Copenhagen Business College, Denmark Natalie Perez Independent Scholar, Denmark

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Abstract This study explored how a higher education educator experienced and made sense of generative artificial intelligence (GenAI) within teaching, peer collaboration, and assessment practices. Using an Interpretative Phenomenological Analysis (IPA) research design, the study focused on a single participant, who is a curriculum leader and instructor at a private European university undergoing rapid shifts in GenAI policy and practice. Data were collected through a semi-structured interview and participant pre-account and analyzed using IPA’s idiographic and interpretative procedures. Findings identified three superordinate themes, including divergent peer adoption of GenAI, pedagogical adaptation and uncertainty, and assessment ambiguity. These illustrate GenAI as an ongoing disruption that reshapes collegial dynamics, teaching approaches, and conceptions of learning and authorship. While GenAI supports accessibility and content simplification, it also raises concerns about student dependency and academic integrity, prompting a shift toward evaluating students’ critical engagement with AI outputs. Interpreted through Mezirow’s transformative learning theory, GenAI emerges as a sustained disorienting dilemma, producing gradual, relational, and ongoing perspective transformation in higher education practice. Future research should extend this work beyond a single-case design to include comparative and longitudinal studies across disciplines and institutions, further examine GenAI-informed assessment models, and the extent to which existing theoretical frameworks capture the relational and evolving nature of AI-mediated educational change. Keywords: GenAI, interpretative phenomenological analysis, higher education, technology

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

The introduction of generative artificial intelligence (GenAI) tools, including large language models capable of producing academic-quality text, synthesizing sources, and generating references on demand, has introduced a disruptive technological shift in higher education (Ajanaku & Thiyagaratnam, 2026; Watson & Rainie, 2025). Unlike earlier digital technologies that augmented existing pedagogical practices, GenAI challenges foundational assumptions about teaching and learning, including what constitutes original student work, how knowledge is produced and validated, and how academic integrity is defined in environments where human and machine generation may increasingly overlap (Balart et al., 2026). As a result, institutions, educators, and students are navigating these disruptions in real time, often without clear or established frameworks to guide decision-making (Sharma & Kumar, 2026). In fact, institutional responses have varied widely, ranging from prohibiting GenAI use to cautious integration and, in some cases, to mandated AI literacy as a graduate competency (Murgatroyd & Couture, 2026). At the same time, individual educators have been confronted with having to interpret and respond to these changes with differing levels of preparation, pedagogical orientation, and institutional support (Chapman et al., 2026). Some students, meanwhile, have adopted GenAI tools at a faster pace than institutional policies or clear conditions for usage have been regulated or established (Ketsman & Lazarevic, 2026). While studies have mapped the prevalence and patterns of GenAI use in higher education (Jamal Eddine et al., 2026), less is known about individual educational experiences, as instructors work to make sense of shifting institutional contexts and teaching expectations (McAllister, 2026). To address this gap, this paper presents an Interpretive Phenomenological Analysis (IPA) of one higher education educator’s experience in teaching a business course during a period of rapid institutional and cultural changes shaped by GenAI adoption, across multiple levels of a higher education institution. The participant occupies a dual role as both a curriculum leader and instructor, positioning them at the intersection of curriculum design, institutional governance, instructional delivery, peer collaboration, and student assessment (Note: GenAI and AI are used interchangeably within this study, and for this study, both concepts refer to content generated by AI models). From a theoretical perspective, this positioning is particularly significant because transformative experiences may occur gradually or rapidly as individuals encounter accumulating pressures or critical events that challenge established ways of working. These experiences often prompt reflection, dialogue with others, and potential shifts in understanding, which may ultimately lead to changes in perception and practice. Consequently, this dual role offers a rich lens for examining how GenAI-related disruption is experienced simultaneously across institutional, relational, and teaching domains. Building on this framing, the study is guided by the following research question: How does a higher education educator experience and make sense of institutional expectations, peer collaboration, teaching, and assessment in the context of widespread GenAI adoption?

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Theoretical Framework This study is grounded in Mezirow’s (1991, 2000) transformative learning theory, which explains how individuals revise and reconstruct their meaning structures in response to experiences that challenge existing assumptions. The theory focuses not only on the acquisition of knowledge or skills, but on how individuals interpret, question, and potentially transform the frames of reference through which they understand themselves, others, and their professional practice. Transformative learning theory is particularly concerned with shifts in meaning that occur when prior expectations or habitual interpretations are no longer sufficient to make sense of experience. In such moments, individuals may begin to critically reassess the assumptions that underpin their beliefs and actions, leading to changes in perspective over time. See Figure 1. Figure 1 Mezirow’s (1991, 2000) Transformative Learning Theory Dimensions

There are several dimensions of Mezirow’s (1991, 2000) transformative learning theory, and they are presented in more detail: 1. Meaning Perspectives and Meaning Scheme For Mezirow, meaning perspectives are broad, often tacit frames of reference that shape how individuals interpret experience, while meaning schemes are the more specific

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

beliefs, values, and assumptions that give these perspectives content (Cranton, 2006; Mezirow, 1991). In higher education, these may include assumptions about authorship, assessment, learning effort, and the role of technology in teaching and learning. 2. Disorienting Dilemma Transformative learning begins with a disorienting dilemma, which is an experience that cannot be adequately understood within existing meaning perspectives and therefore exposes their limitations (Mezirow, 1991, 2000). Such dilemmas may emerge gradually or suddenly (Taylor, 2008). In higher education contexts, they may arise at institutional, educator, or learner levels, often linked to shifts in practice, expectations, governance, or technology use. 3. Critical Reflection and Rational Discourse Following disruption, transformation progresses through critical reflection, particularly premise reflection, which involves questioning the origins and validity of one’s assumptions (Brookfield, 2000; Mezirow, 1991). However, reflection alone is not sufficient, and Mezirow (1991, 2000) emphasizes rational discourse as a complementary process in which assumptions are tested, refined, and negotiated through dialogue with others, such as communities of practice (Taylor, 2008). In higher education, this may occur among colleagues, between educators and students, and within institutional structures. 4. Perspective Transformation Perspective transformation refers to a fundamental shift in how individuals understand themselves and their world (Mezirow, 1991, 2000). Rather than incremental adjustment, it involves a reorganization of meaning perspectives that can result in changes to professional identity, beliefs, and pedagogical practice (Cranton, 2006; Kitchenham, 2008). Overall, transformative learning theory provides a lens for understanding change as a process of meaning reconstruction rather than simple adaptation. It highlights how disruption becomes significant when existing assumptions are no longer sufficient, prompting individuals to critically reflect on, test, and potentially revise their frames of reference. Importantly, these processes may occur unevenly across individuals and contexts, resulting in asynchronous and layered patterns of transformation within higher education environments. Literature Review GenAI in Higher Education Scholars report that GenAI is being used in higher education through student adoption, instructional curriculum, and uncritical integration (Krause et al., 2025). However, concerns arise about the unanticipated impacts of GenAI use, such as the potential to undermine critical thinking and independent learning, which has caused some educators to shift towards more

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

structured and critically engaged integration (Krause et al., 2025; Morris, 2025). While informative, several studies are oriented toward institutional and curricular design, offering guidance on how institutions should respond to GenAI while not covering as much about how the individuals working within these institutions experience such shifts. Relatedly, within higher education, one prevalent concern is academic integrity, as traditional assessment models assume individual authorship, yet GenAI challenges this assumption by generating text that can be indistinguishable from human writing (Zlotnikova et al., 2025). Detection tools further complicate responses due to inconsistent reliability (Sun, 2026), a finding that itself points to a tension between treating authorship verification as a technical problem and recognizing it as an interpretive judgment increasingly left to individual instructors. As a result, assessment scholarship increasingly emphasizes redesign toward authenticity, process-based evaluation, and critical engagement with AI outputs rather than product originality alone (Tran & Dinneen, 2025). However, this redesigning of assessment seems to assume that clear frameworks and guidelines will result in effective implementation, but this assumption is underexplored at the educator level, where instructors must translate policy into everyday teaching practice. GenAI and Educator Experience Research indicates that educator responses to GenAI span a spectrum from enthusiasm to concern, with professional development playing a key role in shaping adoption (Cespedes, 2025; Vivas-Urias et al., 2026). However, this spectrum provides limited insight into how and why educators develop particular orientations toward GenAI (Dishari, 2026). For example, Ellis et al. (2025) found that teachers’ experiences teaching with GenAI differ qualitatively, suggesting that current adoption frameworks may underestimate the complexity and heterogeneity of educators lived engagement with the technology. Beyond technical readiness, GenAI challenges educators’ professional identities as they reconsider their roles in maintaining student creativity, academic integrity, and critical thinking (Sohail et al., 2025). Rather than replacing traditional teaching practices, GenAI is shifting educators’ roles from knowledge transmitters toward facilitators who help students critically evaluate and engage with information (Dochia, 2025). Recent research supports this identity-based perspective, as Ghiasvand and Seyri (2025) found that AI adoption contributes to professional identity reconstruction through changes in expertise and reflective practice. However, the emotional dimensions of GenAI transformation, particularly how educators reconcile professional values with evolving roles, is underexplored beyond a limited number of studies (Dishari, 2026). GenAI and Assessment Recent literature addressing GenAI and assessment has highlighted how educators are shifting their approaches to assessing learning (Heil et al., 2025), with many educators focusing on approaches that prioritize reasoning, iteration, and critical engagement over final outputs (Kofinas et al., 2025). This shift reflects a broader reconceptualization of assessment, reinforcing that traditional forms of evaluative assessment can be manipulated through the use of GenAI and different forms of assessment are needed (Kofinas et al., 2025); this body of work is persuasive on what assessment redesign should look like and why it is ethically

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

necessary, but it is focused on assessment design or broader policies, with limited empirical attention to the labor educators undertake to enact these redesigns and educator’s experiences in seeking to redesign learning assessment in the era of GenAI. For instance, some researchers argue that if GenAI tools are used in assessments, performance could be based on students' knowledge of the tool's competency, rather than disciplinary understanding (Kangwa et al., 2025). This concern is reflected in research identifying assessment anxiety and distrust as key emotional responses among faculty, including suspicion of student work and a shift from evaluating learning to monitoring authorship (Dishari, 2026). These findings suggest that assessment reform is not only a matter of redesigning assessment tools but also involves a personal and professional adjustment for educators responsible for implementation. When examining the literature, a common theme is found, which is that GenAI's disruption to higher education is increasingly documented at the level of institutions, curricular design, and assessment, with some literature addressing educator identity and affect. What remains underexplored is the experiential and developmental dimension of that disruption, such as how individual educators move through identity renegotiation and practice transformation over time, rather than where they land on a spectrum. This study addresses that gap by centering educator experience as a process to be examined. Methodology Research Design This study uses an IPA research design. IPA is a qualitative approach that aims to explore how individuals make sense of their lived experience, with a particular focus on experiential meaning-making in contexts of personal significance and complexity (Smith et al., 2009). IPA is grounded in phenomenological philosophy and is committed to a detailed, empathic engagement with participants' accounts of their lived experience (Smith et al., 2009). It is particularly well-suited to the present study because it does not seek to generalize across populations but to illuminate the particularity of individual experience in depth (Smith et al., 2009). One important feature of IPA is the double hermeneutic, where the researcher interprets a participant who is themselves engaged in interpreting their own experience (Smith et al., 2009). This reflexive, interpretative dimension is especially pertinent to the present study, in which the participant is not simply describing external events but actively constructing meaning in relation to a rapidly changing educational landscape. Participant Selection IPA research typically leverages purposive sampling with a small number of participants, prioritizing depth of analysis over breadth of coverage (Smith et al., 2009). The present study is based on a single participant, which is an approach consistent with the idiographic commitments of IPA and particularly appropriate when the aim is to produce a rich, theoretically grounded account of a specific experiential configuration (Perez, 2022). The participant was selected based on their dual role as both module leader and instructor in a final-

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

year undergraduate business strategy course, a positioning that afforded access to GenAIrelated disruptions across multiple relational and institutional domains simultaneously. Participant Details The participant, Thomas (pseudonym), is a male educator aged 40-50. He works at a private university in Europe. Thomas has approximately 15 years of experience teaching higher education and holds a dual role as both a curriculum leader and instructor, primarily teaching 3rd- and 4th-year undergraduate students. In this capacity, Thomas is in a complex position at the intersection of curriculum leadership, peer instructional collaboration, and student assessment within an institutional environment that has, only recently, integrated GenAI into teaching and learning. This dual role situates Thomas within multiple overlapping spheres of influence, requiring continuous negotiation between institutional expectations, disciplinary norms, and classroom-level pedagogical decision-making. Importantly, Thomas’s experiences highlight a radical institutional shift in views and practices towards GenAI. At the beginning of last year, his institution viewed GenAI as dangerous; those views changed towards the end of last year, as GenAI was viewed as distracting. However, this year, Thomas’s institution has changed its views on GenAI, now considering it essential. From an interpretative phenomenological perspective, Thomas’s lived experience reflects how GenAI is not merely a technological tool but a disruptive force reshaping teaching norms, assessment practices, and professional expectations within his higher education institution. For the purposes of this study, we asked Thomas for a pre-account of his experiences with GenAI before the interview. He provided more context in his pre-account across three areas. Instrument A semi-structured, in-depth interview was used for this IPA study. The interview was designed to facilitate idiographic, reflective, and experiential accounts, allowing the participant to describe lived experience in their own terms while enabling flexible probing for depth and meaning making. The participant provided informed consent prior to the interview and was advised of their right to withdraw at any point. All identifying information has been removed or altered to protect participant anonymity (see Appendix for questions). Analytical Process This study followed the IPA procedure described by Smith et al. (2009) to analyze the data. The transcript was read multiple times to develop familiarity with the data. Initial notes were made, capturing descriptive, linguistic, and conceptual observations. Emergent themes were then identified, representing the researcher's interpretative engagement with the participant's meaning-making. Themes were subsequently clustered into superordinate themes that captured patterns of experience across the transcript. Throughout this process, the researcher maintained a reflexive awareness of the interpretative nature of the analysis and of the theoretical framework's influence on the interpretative process, a requirement of the double hermeneutic. Importantly, the theoretical framework was used to support and guide the instrument creation,

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

but it was not used to analyze the data; instead, the framework was applied to the data, after the fact, to determine whether or not the participants’ experiences aligned or misaligned with the theory (Perez, 2022). Researchers’ Positionality and Reflexivity Statement In IPA, the researcher is not a detached, objective observer, but an active interpretative tool coconstructing meaning with the participant through a double hermeneutic. Because complete bracketing of researcher assumptions is inconsistent with IPA epistemology, the authors disclose their own personal and professional positionalities to clarify the reflexive lenses that shaped data collection, analysis, and thematic interpretations. Author 1 (Karen), who conducted the semi-structured interviews, is a multicultural, multilingual researcher who has lived, studied, and taught in the U.S. states of Washington, Florida, California, Arizona, and Hawaii, as well as in Japan, the United Kingdom, and Denmark. Her firsthand familiarity with the European and Danish higher education systems, combined with her experience navigating diverse organizational shifts, allowed her to establish a close, empathetic rapport with Thomas as he navigated his own institutional upheavals. Rather than seeking an artificial, detached ‘objectivity’, Karen practiced continuous reflexivity, recognizing that her own experiences as an educator helped her examine the emotional nuances of Thomas’s professional isolation and pedagogical uncertainty, while simultaneously using an interview protocol to remain firmly grounded in his individual, idiographic narrative. In short, Karen used a semi-structured interview protocol to ensure that participant voice directed the dialogue, while actively monitoring how her own background as an international educator influenced her listening, prompting, and interpretive focus during data collection. Author 2 (Natalie) served as the primary data analyst, bringing over 15 years of experience in higher education and organizational psychology with an educational background spanning Hawaii, France, and China. Her psychological background directly informed her interpretation of how Thomas cognitively and emotionally coped with, adapted to, or resisted changes to his professional identity. Additionally, Natalie’s technical expertise in qualitative design, machine learning, and natural language processing (NLP/NLG) methods shaped how she interpreted Thomas's descriptions of GenAI tools. While her technical literacy allowed her to deeply grasp the mechanics of the ‘authorship to editorship’ shift Thomas described, she reflexively questioned whether her own comfort with technology could lead her to misrepresent Thomas's anxieties about students “cheating their own education”. Overall, through active dialogue and a shared commitment to the double hermeneutic, the authors sought to co-construct an interpretation respecting the depth and emotional complexity of Thomas’s lived reality.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Results The analysis identified three superordinate themes that capture how the participant experiences and makes sense of teaching, assessment, and collaboration in the context of widespread GenAI disruption within their higher education institution (see Table 1 for an overview). Each theme reflects a distinct dimension of the participant’s lived experience and is supported by at least three statements to support inclusion (Smith et al., 2009). The themes are presented in the following sections. Table 1 Superordinate and Subordinate Thematic Findings Superordinate Themes

Subordinate Themes

Superordinate Theme 1. Negotiating Divergent GenAI Adoption Among Peers

Divided AI-Use

Superordinate Theme 2. Pedagogical Adaptation and Intellectual Uncertainty with GenAI

GenAI as Pedagogical Rupture

Superordinate Theme 3. Assessment Ambiguity

GenAI Use as Uncontrollable

Hidden vs. Open AI-Use and Social Judgment Tight, Relational Team

Decomplexifying Learning Materials Shifting from Authorship to Editorship

Assessment as Analysis Avoiding Cheating in Education

Superordinate Theme 1: Negotiating Divergent GenAI Adoption Among Peers This superordinate theme contains three subordinate themes, which are as follows: Divided AI-Use Thomas describes his teaching cohort as both collaborative and pedagogically divided in their engagement with GenAI. Although Thomas shared that his teaching cohort collectively acknowledges that AI has become embedded in contemporary student learning practices, his teaching cohort remains divided on the ways his colleagues understand, adopt, and disclose AI use in their teaching. For instance, Thomas shared, “We all acknowledge that AI is now a significant part of students’ research.” However, he went on to say that he and his peers “use it [GenAI] in different ways.” Within his teaching team, Thomas explained that three of his

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

peers “do not use it [GenAI] at all,” while another uses it selectively, and yet another “exclusively uses AI” in their teaching. His peer, who is a GenAI adopter, uses AI for almost everything, including “class exercises, class slides,” and openly shares AI-generated materials with the teaching team. Interestingly, Thomas’s accounts suggest that these differences are interpreted through broader assumptions about technological competence and domain expertise. He describes his colleagues who resist AI as “old-school” and goes on to explain that he does not believe they have the “technological skills needed to use AI.” In this sense, Thomas seems not to be applying a generational label to the concept of “old-school,” but rather a technological literacy label to his peers who have not adopted AI; perhaps not surprisingly, Thomas shares that these same colleagues do not indicate any willingness to upskill or build their technological skills. It is this seeming rejection of AI that suggests they are more “traditionalists” and that they teach in ways that are increasingly viewed as canonical to Thomas. Hidden vs. Open AI-Use In a different instance, Thomas explains that one of his other teaching peers actually uses GenAI to strengthen their subject understanding, but they have not disclosed this practice with their colleagues, while a different colleague actively and openly uses GenAI in their teaching practices. Thomas interprets this unwillingness to share as the potential for social judgment, explaining that he believes his colleague might feel "embarrassment or have low self-esteem,” as they have not been “formally educated or trained to teach.” This unwillingness to share GenAI use suggests that, when engaging with non-adopters, some GenAI-adopters might choose to hide their AI use, as they may view it as influencing perceptions of social judgment or illegitimacy within academic environments. However, not all GenAI adopters may experience social judgment within academic environments, as observed through Thomas’s own experiences and those of a vocal peer who uses GenAI openly. For example, Thomas himself had adopted GenAI for tasks outside of his classroom, and it wasn’t until later in the semester that he realized he could use GenAI to create more class exercises, expand current exercises, and create more simulated activities to aid learners. By the end of the semester, Thomas himself actively and openly used GenAI. He views AI use in the classroom as an important tool and values GenAI for helping him be more flexible, creating content that works best for his learners. Alternatively, one of his peer instructors has positioned themselves as an active and open AI user. Initially, in conversations with this individual, Thomas noted that this peer resisted GenAI use, but they later decided to experiment with AI and found it helpful and useful. Not long after, this high adopter peer started using GenAI during classroom activities, particularly in response to student requests for additional learning support. However, this peer's attempts to share AI-generated materials with colleagues revealed varying levels of openness toward AIuse and change. Thomas noted that one colleague simply rejected the immediate implementation of AI-generated materials because they felt it was “too late” to modify existing teaching plans; although they expressed willingness to use the materials in a future semester. These experiences reflect how the mere use or reference of AI can be off-putting for some instructors; however, this was not the case for all. Timing seemed to be a key factor for one instructor considering the adoption of AI-generated materials.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Tight, Relational Team Despite his peers’ differing views of GenAI use, Thomas repeatedly emphasizes that he feels his teaching cohort is collaborative and emotionally supportive. Although Thomas has experienced tensions and “bumps on the road” with his cohort based on differing AI use and views, Thomas believes that his teaching team ultimately has developed a sense of cohesion and trust. He explains it this way, “We have become a tight little team,” and “We have found a sweet spot” in working together. This sweet spot seems to center on respect and collegiality; although Thomas and his peers don’t always see eye to eye, being respectful, despite their differences, has been an important part of fostering relationships among them. And Thomas’s reflections suggest that these collegial relationships function as an important source of stability amid technological and institutional uncertainty. He described feeling that his team of teachers was “trustworthy and responsible.” Having become a “tight team” amid the tumultuous semester, where their institution changed gears and decided to support student GenAI use in the classroom, Thomas feels that he is going to “miss this team.” This reinforces that having a support group, even for those who are not adopters, during a technological change, is valuable for educators. From Thomas’s perspective, human relationships remained one of the most stabilizing and meaningful aspects of navigating pedagogical change in the GenAI era. Superordinate Theme 2: Pedagogical Adaptation and Intellectual Uncertainty with GenAI This superordinate theme contains three subordinate themes, which are as follows: Pedagogical Rupture Thomas describes GenAI as both a pedagogical rupture and a requirement for changes to his teaching. When considering his views on GenAI in the classroom, Thomas reflects on his experiences within the context of his institution’s shifting GenAI norms. He explains that his college initially viewed AI as “a dangerous tool for students to use,” particularly because of concerns surrounding academic integrity and student overreliance on AI-generated content. However, as AI has continued to grow in external popularity, his institution switched gears, viewing AI as more of a distraction and less of a danger. In the same way, Thomas shared that he observed his students increasingly use or at least submit work that looked to be GenAIcreated. It was the repeated instances of submitted student work that looked like it was created by GenAI that Thomas realized he had to shift from GenAI resistance towards adaptation. Reflecting on this transformative experience, he realized that he had to “embrace AI and educate students on how to use it appropriately.” However, this transition was not straightforward; it involved continual exploration. Decomplexifying Learning Materials Thomas explained that he values GenAI as a pedagogical support tool, but it took time for him to realize that it could be a support tool to improve "accessibility and classroom engagement."

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Thomas spent time experimenting with AI-generated teaching materials, and through his experimentation, felt “positive” about using it, as he felt that AI did a good job at presenting concepts in “layperson’s terms" rather than an academic style.” This communicative change was important to Thomas, as he felt like using AI to translate complex academic language into more simplistic communication made learning more approachable for students. From this perspective, Thomas’s GenAI adoption took time and involved experimentation, as he engaged with, examined, and explored GenAI capabilities. Through that experimentation, Thomas found that complex-language translation was one of the most valuable ways to use GenAI to simplify the learning experience. In this way, Thomas values GenAI for its communication and translation abilities, shifting from the language of the academy to language that makes the most sense for his learners. Shifting from Authorship to Editorship At the same time, Thomas expresses significant concern regarding students’ growing dependence on AI-generated content. He describes observing classroom practices, noting that “98% of the class uses AI already,” and describes many student activities as becoming “all AIgenerated.” These experiences challenge his assumptions about intellectual effort, authenticity, and meaningful engagement with learning. Thomas explains that he recognizes that AI use has become unavoidable within higher education, but he remains concerned that students will rely on AI for “copy/paste responses” without engaging critically with the material. He finds this tension difficult to navigate, as, on one hand, he appreciates AI’s pedagogical utility while simultaneously fearing its potential to diminish critical thinking and independent learning. He holds these perspectives concurrently, as he considers how GenAI is shaping students' learning, thinking, and engagement with knowledge and authorship. However, more recently, Thomas’s views have continued to shift, as he said, “We must embrace AI, follow the trends, stay relevant, and help our students understand good from bad AI.” This is an interesting perspective and suggests that there are still good ways to use AI and problematic ones. Digging deeper, Thomas believes that “good” AI use means knowing how to use AI appropriately: creating outputs that directly answer assignments, ensuring they are accurate, and knowing what good writing looks like before submitting GenAI content. These reflections indicate a shift from original authorship to that of an editor; rather than creating and editing, learners can skip the creation and focus on creating strong editing skills. Nevertheless, this still begs the question of whether an editor, rather than an author role, diminishes critical thinking or simply shifts learner tasks. Superordinate Theme 3: Assessment Ambiguity This superordinate theme contains three subordinate themes, which are as follows: GenAI Use as Uncontrollable Thomas describes assessment in the GenAI era as increasingly unstable and difficult to regulate, particularly as AI-generated content becomes embedded within nearly all aspects of

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

student learning. Reflecting on the scale of AI use among students, Thomas explains that almost all his students use GenAI these days. He reports that his students’ work extends beyond written coursework, noting that “Everything is AI-assisted, even their oral group presentations.” For Thomas, the use of AI has fundamentally disrupted traditional views on authorship and the authentic demonstration of student knowledge. He repeatedly acknowledges that AI use has become unavoidable within his classroom and his higher education institution, explaining, “We can’t control this environment.” Although he mentioned that strategies such as handwritten assessments or restricted internet access could limit AI use, Thomas believes students will continue to find ways around institutional controls. To this end, he suggests that AI use in assessment is inevitable, suggesting that rather than push upstream, he has decided to accept GenAI use in assessments. Assessment as Analysis Thomas’s views on what to assess and on the quality of work have changed with his support for AI adoption. Rather than viewing AI use itself as inherently problematic, Thomas has increasingly focused on students’ ability to critically analyze, interpret, and defend AIgenerated outputs. He explains, “It is not about telling us, but analyzing what AI provided them,” suggesting that assessment, for Thomas, is becoming less about whether AI was used and more focused on how students intellectually engage with AI-generated information. Similarly, when discussing grading quality, Thomas states, “It depends on how the student(s) can analyze what they copied from AI,” emphasizing critical thinking, justification, and the use of credible supporting sources as central markers of acceptable work. These findings indicate a new view of assessment that aims less to examine authorship or one’s ability to articulate their learning, and more about knowing what a good output looks like by critically examining outputs and making final decisions on what should be submitted or not. From this lens, assessment is more about one’s ability to analyze and edit than to create. Avoiding Cheating in Education Although Thomas views GenAI for learning and teaching in a positive light, he also has some reservations about fair and ethical assessment. He reflects on his early teaching experiences, when GenAI was not as popular, and he lacked a heightened awareness of AI.” Thomas admitted that ‘some students’ marks were higher than they should have been due to my oversight with AI.” In reflection, he believes he should not have given his learners as high marks as he did, as he believed their work was original and authentically created by them, rather than machine-generated. These reflections suggest that perhaps Thomas feels he must raise the quality bar, expecting higher-quality outputs with GenAI than in instances where GenAI has not been used. He goes on to suggest that his evolving awareness of GenAI makes traditional assessment practices no longer viable. Thomas also expresses frustration that many of his students appear more interested in “shortcuts and superficial information” than in meaningful intellectual development, leading him to feel that some students are “cheating their own education.” Yet, despite these concerns, he does not advocate for blanket punishment or prohibition of GenAI. Instead, Thomas increasingly positions his role as helping students learn

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

how to critically evaluate AI-generated content responsibly and ethically. Thomas feels that by supporting learners in knowing how to critically evaluate GenAI, he can support their education and mitigate shortcuts they feel willing to make but does not. Discussion This study explored how one higher education instructor, Thomas, navigated the rapid emergence of GenAI within his teaching practice, collegial relationships, and assessment approaches. Thomas’s experience indicates that technological transformation occurred incrementally through repeated encounters with GenAI adoption, institutional change, and shifting pedagogical expectations, supporting Cranton and Taylor’s (2011) view that disorienting dilemmas may emerge gradually. However, his case also highlights an important distinction between adaptation, accommodation, and perspective transformation. Adaptation involves adjusting behaviors or practices in response to changing conditions, whereas accommodation involves integrating new information while maintaining existing assumptions (Mezirow, 1991). In contrast, perspective transformation represents a fundamental shift in meaning perspectives, assumptions, or professional identity through critical reflection and revised frames of reference (Mezirow, 1991, 2000). Thomas’s experiences demonstrate movement across these processes but raise questions about whether his evolving GenAI practices represent full perspective transformation or an ongoing process of adaptation and accommodation. Implications for Practice Social Negotiation of GenAI Use Among Educators The first superordinate theme demonstrates that GenAI adoption within teaching teams can be uneven, shaped by varying levels of technological confidence, pedagogical beliefs, and openness to change. Thomas described colleagues ranging from complete non-adopters to highly active AI users, reflecting how educators can occupy varying positions along an adoption continuum. Importantly, resistance to AI was interpreted less as generational and more as linked to technological literacy and willingness to upskill. The findings also reveal that some educators may conceal their GenAI use due to fears of social judgment or perceived illegitimacy, suggesting that AI use remains socially sensitive within academic environments. Despite these differences, Thomas consistently emphasized the importance of collegial trust and collaboration, highlighting how supportive peer relationships can provide stability during periods of technological uncertainty. GenAI as Democratizing Information Access and Complicating Evidence of Learning The second superordinate theme illustrates how GenAI disrupted Thomas’s assumptions about teaching and learning. Initially resistant due to concerns about academic integrity, Thomas gradually shifted toward adaptation as student use of AI became increasingly common. One key benefit he identified was GenAI’s ability to simplify complex academic language into more

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

accessible forms, supporting student engagement and accessibility. However, Thomas also expressed concern that students were becoming overly reliant on AI-generated outputs. His reflections suggest a shift from traditional notions of authorship toward a model in which students increasingly act as editors, analysts, or evaluators of AI-generated content. While Thomas viewed critical evaluation of GenAI outputs as an important emerging skill, he remained uncertain whether this shift supports or diminishes critical thinking and authentic learning. Reimagining Assessment in GenAI-Mediated Learning The third superordinate theme highlights growing uncertainty surrounding assessment and academic integrity in AI-mediated learning environments. Thomas described GenAI use as largely unavoidable and difficult to regulate, leading him to question the viability of traditional anti-cheating approaches. Rather than focusing on whether students used GenAI, Thomas increasingly prioritized students’ ability to critically analyze, justify, and refine AI-generated information. In this sense, assessment became less focused on original authorship and more focused on evaluative reasoning and analytical engagement. At the same time, Thomas worried that some students were using AI as a shortcut, “cheating their own education” by avoiding deeper intellectual engagement. Nevertheless, he did not advocate banning GenAI; instead, he positioned ethical and critical GenAI use as a key educational responsibility. Implications for Theory This study extends Mezirow’s (1991, 2000) transformative learning theory by examining where Thomas’s experience with GenAI aligns with and departs from the theory. Consistent with Cranton and Taylor’s (2011) view that disorienting dilemmas may emerge gradually, Thomas’s experience reflects an extended process of disruption through changing institutional expectations, student GenAI use, and evolving pedagogical practices. His reflections suggest elements of transformative learning, particularly through critical questioning of assumptions about authorship, assessment, and teaching. However, his experience also reveals tensions within the framework, as his responses sometimes seem to reflect adaptation and accommodation rather than a complete perspective transformation. Overall, Thomas’s case illustrates explanatory value and limitations of Mezirow’s theory in understanding how gradual technological change occurs in a higher education setting. Meaning Perspectives and Meaning Scheme Initially, Thomas’s meaning perspectives and meaning schemes were grounded in traditional assumptions about higher education, where authentic learning was associated with independent student effort, original authorship, and critical engagement with course content. He also operated within institutional narratives that framed GenAI as potentially “dangerous” due to concerns surrounding academic integrity and overreliance on AI-generated work. These assumptions seemed to have shaped his initial understanding of GenAI’s role in education, assessment, and knowledge production. At the same time, Thomas also valued adaptability and

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

professional relevance, creating underlying tension between preserving traditional educational norms and responding to changing technological realities. Disorienting Dilemma Thomas’s disorienting dilemma emerged gradually through repeated exposure to GenAI in his classroom. Rather than a single transformative moment, his disruption resulted from ongoing encounters with students' use of AI, institutional policy shifts, and collegial differences surrounding GenAI adoption. Observing that most students were already using AI, including in written work and oral presentations, challenged his assumptions about authenticity, controllability, and assessment validity. Simultaneously, his institution shifted from framing GenAI as a threat to accepting its inevitability within higher education. These experiences seemed to have destabilized Thomas’s prior understanding of teaching and learning because traditional assumptions about authorship, originality, and assessment no longer adequately explained the educational realities he was experiencing. It is worth noting that the disruption Thomas describes here is difficult to disentangle from his institutional compliance, as Thomas's institution moved from prohibition to acceptance of GenAI. And it is difficult to confidently assert whether transformation truly took place, as Thomas was adjusting to an environment that had instituted a new policy, independent of any internal adjustments to his own assumptions. Critical Reflection and Rational Discourse Following this disruption, Thomas engaged in critical reflection and rational discourse. Through critical reflection, he questioned whether resisting GenAI remained realistic, whether assessment should continue prioritizing original authorship, and whether meaningful learning might increasingly involve evaluating and refining AI-generated outputs rather than solely producing original content. His reflections demonstrate premise reflection, as he reconsidered deeply held assumptions about academic integrity, intellectual effort, and educational value. Rational discourse also played a key role in shaping his transformation. Conversations with colleagues exposed him to differing perspectives on GenAI adoption, ranging from resistance to extensive integration. Collaborative discussions, experimentation, and collegial negotiation helped Thomas test and refine his assumptions about the use of GenAI in teaching. Importantly, his experiences suggest that this change in thinking was not solely cognitive, but also relational and emotional, shaped by collegial trust, fears of judgment, and supportive peer relationships. That said, we can also view Thomas’s experiences in light of his exposure to a range of collegial positions about AI, and the mere exposure to different viewpoints could be viewed, in a different light, as social learning and perhaps some behavioral recalibration, as he adjusts to different views about GenAI, while attempting to retain his positionality as the curriculum lead.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Perspective Transformation, Accommodation, or Adaptation Over time, Thomas demonstrated evidence of perspective transformation through substantial changes in how he understood GenAI’s role in education. He moved from initial resistance toward a more adaptive position that accepted GenAI as unavoidable within higher education. Rather than focusing exclusively on preventing GenAI use, Thomas increasingly emphasized helping students critically evaluate and ethically engage with AI-generated content. His understanding of assessment also shifted, moving away from authorship detection and toward evaluating students’ analytical and editing abilities. In this sense, Thomas reconceptualized educational competence as involving the ability to critique, refine, and justify AI-assisted outputs rather than simply producing entirely independent work. Although he continued to express concerns about superficial learning and students “cheating their own education,” his overall frame of reference shifted toward adaptation rather than prohibition. However, while some of Thomas’s behavior suggests transformation, others do not. For example, some of what Thomas described seems more aligned with adaptation. This occurs, for instance, during his acceptance that GenAI use among students is unavoidable, and his shift away from detection-focused enforcement, reflects a practical adjustment, based on conditions he could not control. Alternatively, in a different situation, Thomas suggests more of an accommodation than transformation, as Thomas continues to value original authorship and continues to worry about students “cheating their own education,” even as he built new assessment practices around evaluating and refining AI-assisted work. While some of Thomas's account does support a perspective transformation, specifically his reconceptualization of educational competence as involving the ability to critique, refine, and justify GenAI-assisted outputs, and his shift in what he takes assessment to be measuring. These changes are at tension with other parts of Thomas’s account, which suggests he might experience partial or ongoing transformation. In this way, Thomas's continued ambivalence about “cheating their own education” indicates that his old and new meaning perspectives somewhat coexist rather than one having fully displaced the other. Tensions with Transformative Learning Analyzing Thomas’s experience through Mezirow’s stages illustrates some tension. First, not all changes he described represent transformative learning. Thomas’s growing ability to evaluate GenAI-generated content reflects skill acquisition, which might take place through transformation, but it does not explicitly indicate a shift in his underlying assumptions. In addition, some of his adaptations occur in-step with his institutional expectations, as Thomas adopted GenAI-related practices as changes were occurring, which might not suggest transformation, but merely adaptation. Lastly, Thomas’s continued concerns about students “cheating their own education” indicate a greater tension between adaptation and transformation, which begs the question whether Thomas’s responses highlight how he coped, rather than a fully integrated new perspective. From these instances, GenAI, for Thomas, may represent a persistent rather than discrete disruption, which might be difficult for educators to

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

reach a stable transformed perspective. In all, Thomas’s case might be best understood as a layered process involving skill acquisition, adaptation, accommodation, coping, and partial transformation. Limitations This study contains several limitations that should be considered when interpreting the findings. For one, the study used a single-participant IPA design, meaning the findings reflect the lived experiences and interpretations of one educator situated within a specific institutional and disciplinary context. While IPA intentionally prioritizes depth, idiographic analysis, and experiential richness over breadth and statistical generalizability (Smith et al., 2009), the findings cannot be assumed to represent the experiences of all higher education educators navigating GenAI or institutional disruption with respect to GenAI. Thomas’s experiences were shaped by his unique dual role as both a curriculum leader, supporting a teaching cohort of peers, and an instructor at a private European university, and educators (working in different institutional, cultural, disciplinary, or governance contexts) may experience GenAI-related disruptions differently. In addition, the study relied exclusively on self-reported data and participant pre-accounts, which means the findings are inherently interpretative and mediated through Thomas’s retrospective meaning-making. As with all qualitative self-report methods, there remains the possibility of selective recall, exaggeration, omission, or social desirability influencing how experiences, colleagues, students, and institutional practices were represented, particularly given the sensitive and contested nature of GenAI use, academic integrity, and professional competence in higher education (Creswell, 2018). Additionally, the study captured experiences during a rapidly evolving technological and institutional period, in which institutional policies, pedagogical norms, and GenAI capabilities continue to shift rapidly. Consequently, the findings should be understood as temporally situated in an early, transitional phase of GenAI integration rather than stable or enduring representations of educational practice. Although IPA recognizes the importance of the double hermeneutic, in which the researcher interprets the participant’s own meaning-making, the interpretative nature of the analysis introduces the possibility that alternative researchers may have emphasized different experiential dimensions, thematic structures, or theoretical interpretations. While reflexivity was maintained throughout the analytical process, complete bracketing of researcher assumptions is neither fully possible nor consistent with IPA epistemology (Smith et al., 2009). The use of transformative learning theory as a post-analytical interpretative lens may have foregrounded processes of disruption, reflection, and transformation while potentially backgrounding other important dimensions of GenAI adoption, such as power relations, institutional governance, emotional labor, digital inequality, or sociomaterial entanglements that alternative theoretical frameworks may have illuminated more explicitly. Relatedly, the study focused solely on an educator perspective and did not incorporate student, administrative, or policy-maker voices, limiting the ability to examine how GenAI-related transformations are negotiated relationally across broader institutional ecosystems. Additionally, because the

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

findings are grounded in one participant’s perceptions of colleagues and students, the study cannot independently verify the extent or nature of others’ AI use, beliefs, or behaviors. Despite these limitations, we believe the study offers a rich, theoretically informed, and contextually grounded exploration of how one educator experienced pedagogical disruption, collegial negotiation, and shifting understandings of learning and assessment during a period of significant technological transformation within higher education. Future Research Recommendations Future research should extend this study beyond a single-case design to explore whether similar experiences of pedagogical disruption, assessment ambiguity, and shifting notions of authorship are evident across disciplines, institutional types, and cultural contexts. We see value in expanding to multi-participant and comparative studies to help establish whether these findings are idiosyncratic or more broadly representative of higher education responses to GenAI. In addition, longitudinal research should be explored to examine how educator perspectives evolve over time as GenAI tools, institutional policies, and assessment practices continue to develop. This would clarify whether perspective transformation stabilizes or remains in a state of ongoing negotiation. In addition, future work should incorporate multiple stakeholder perspectives, particularly students and institutional leaders, to better understand how GenAI-related practices and expectations are interpreted across the education ecosystem. Lastly, we recommend that theoretical and applied research be conducted to test and refine emerging GenAI-aware assessment models, as well as to examine whether existing frameworks of transformative learning fully capture the relational, emotional, and technologically mediated nature of pedagogical change. Conclusion This study illustrates how GenAI is reshaping higher education beyond the adoption of a new technological tool, creating an ongoing process of pedagogical, professional, and institutional renegotiation. For educators like Thomas, GenAI introduces tensions between innovation, authenticity, and academic integrity, requiring continual reflection on what constitutes meaningful teaching and learning. The findings suggest that educator responses may involve a complex interplay of adaptation, accommodation, and emerging transformation rather than a straightforward transition toward AI adoption. For institutions, this highlights the importance of moving beyond restrictive GenAI policies towards clearer pedagogical guidance, sustained professional development, and collaborative environments that support responsible experimentation. Perhaps most importantly, the study suggests that assessment is undergoing a fundamental shift, from evaluating individual production and authorship toward assessing students’ abilities to critically evaluate, refine, justify, and ethically engage with GenAIsupported knowledge. Rather than positioning GenAI as a threat to academic integrity, the findings indicate that higher education institutions may need to reconceptualize what constitutes meaningful learning, intellectual effort, and educational competence in GenAImediated environments.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Declaration of Generative AI and AI-assisted Technologies The authors would like to acknowledge the use of artificial intelligence (AI) tools in the preparation of this manuscript. The following tool was used for this specific purpose only: Grammarly. All AI-edited text was thoroughly reviewed and revised by the authors to ensure accuracy, clarity, and adherence. AI tools were not used for other purposes, including data generation, data analysis, methodology, or interpretation of findings or conclusions. The authors take full responsibility for all aspects of the final manuscript.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

References Ajanaku, O. J., & Thiyagaratnam, E. (2026). Generative AI: a disruptive game changer in higher education. In AI-Driven Revolution: Transforming the Business Landscape, pp. 169–196. https://doi.org/10.1142/9781800616578_0009 Balart, T., Díaz, B., & Shryock, K. (2026). A systematic literature review on the pedagogical implications and impact of GenAI on students’ critical thinking. Algorithms, 19(3), 179. https://doi.org/10.3390/a19030179 Brookfield, S. D. (2000). Transformative learning as ideology critique. In J. Mezirow & Associates (Eds.), Learning as transformation: Critical perspectives on a theory in progress, pp. 125–148. Jossey-Bass. Cespedes, A. A. (2025). Pedagogical shifts in the age of GenAI: faculty perspectives from a higher education context. Journal of Pedagogical Sociology and Psychology, 7(3), 35–48. https://doi.org/10.33902/jpsp.202537244 Chapman, B. L., Ross, A. S., & Petraki, E. (2026). Teacher readiness for generative AI: a theory of planned behaviour approach. Social Sciences & Humanities Open, 13, 102351. https://doi.org/10.1016/j.ssaho.2025.102351 Cranton, P., & Taylor, E. W. (2011). Transformative learning. In The Routledge International Handbook of Learning, pp. 214–223. Routledge. Cranton, P. (2006). Understanding and promoting transformative learning: a guide for educators of adults (2nd ed.). Jossey-Bass. Creswell, J. W. (2018). Research design: Qualitative, quantitative, and mixed methods approaches (5th ed.). Sage. Dochia, I. (2025). The changing role of teachers in the digital age: from knowledge transmitter to learning facilitator. International Journal of Social and Educational Innovation. (IJSEIro), 12(23), 109–118. Dishari, S. (2026). Teaching in the age of generative AI: a qualitative analysis of faculty identity, emotion, and assessment concerns. Journal of Artificial Intelligence and Technology. https://doi.org/10.37965/jait.2025.0941 Ellis, R., Han, F., & Cook, H. (2025). Qualitatively different teacher experiences of teaching with generative artificial intelligence. International Journal of Educational Technology in Higher Education, 22(1), 33. https://doi.org/10.1186/s41239-025-00532-2 Ghiasvand, F., & Seyri, H. (2025). A collaborative reflection on the synergy of Artificial Intelligence (AI) and language teacher identity reconstruction. Teaching and Teacher Education, 160, 105022. https://doi.org/10.1016/j.tate.2025.105022

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Heil, J., Ifenthaler, D., Cooper, M., Mascia, M. L., Conti, R., & Penna, M. P. (2025). Students’ perceived impact of GenAI tools on learning and assessment in higher education: the role of individual AI competence. Smart Learning Environments, 12(1), 37. https://doi.org/10.1186/s40561-025-00395-0 Jamal Eddine, R., Gide, E., & Al-Sabbagh, A. (2026). Systematised evidence mapping of generative artificial intelligence (GenAI) and digital divide phenomena in higher education. Discover Computing, 29(1), 157. https://doi.org/10.1007/s10791-026-10044-w Kangwa, D., Msafiri, M. M., & Fute, A. (2025). Balancing innovation and ethics: promote academic integrity through support and effective use of GenAI tools in higher education. AI and Ethics, 5(4), 3497–3530. https://doi.org/10.1007/s43681-025-00689-6 Ketsman, O., & Lazarevic, B. (2026). Understanding GenAI adoption among undergraduate learners: perceptions, use, and familiarity. Journal of Educational Technology Systems. https://doi.org/10.1080/00472395.2026.1416160 Kitchenham, A. (2008). The evolution of John Mezirow’s transformative learning theory. Journal of Transformative Education, 6(2), 104–123. https://doi.org/10.1177/1541344608322678 Kofinas, A. K., Tsay, C. H. H., & Pike, D. (2025). The impact of generative AI on academic integrity of authentic assessments within a higher education context. British Journal of Educational Technology, 56(6), 2522–2549. https://doi.org/10.1111/bjet.13585 Krause, S., Panchal, B. H., & Ubhe, N. (2025). Evolution of learning: assessing the transformative impact of generative AI on higher education. Frontiers of Digital Education, 2(2), 21. https://doi.org/10.1007/s44366-025-0058-7 McAllister, J. S. (2026). Understanding K-12 Public High School Teachers’ Perceptions of Artificial Intelligence in Education: A Phenomenological Study. Doctoral Dissertations and Projects. 7923. https://digitalcommons.liberty.edu/doctoral/7923 Mezirow, J. (1991). Transformative dimensions of adult learning, (vol. 350, pp. 94104– 1310). San Francisco, CA: Jossey-Bass. Mezirow, J. (2000). Learning to think like an adult: Core concepts of transformation theory. In J. Mezirow & Associates (Eds.), Learning as transformation: critical perspectives on a theory in progress, (pp. 3–33). Jossey-Bass. Morris, D. L. (2025). Rethinking science education practices: Shifting from investigationcentric to comprehensive inquiry-based instruction. Education Sciences, 15(1), 73. https://doi.org/10.3390/educsci15010073 Murgatroyd, S., & Couture, J. C. (2026). AI unplugged: the hype and hope in education futures. Taylor & Francis.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Perez, N. (2022). “Who am I?”: An interpretative phenomenological analysis of tutors’ lived experiences providing online writing tutoring (Order No. 29209402). Available from ProQuest Dissertations & Theses Global. (2691520523). https://www.proquest.com/dissertations-theses/who-am-i-interpretativephenomenological-analysis/docview/2691520523/se-2. Sharma, S., & Kumar, R. (2026). Navigating GenAI in teacher education: an empathetic, verstehen-informed framework. Brock University. https://hdl.handle.net/10464/20018 Smith, J. A., Flowers, P., & Larkin, M. (2009). Interpretative phenomenological analysis: theory, method and research. Sage. Sohail, S., Parveen, S., & Dar, T. (2025). Investigating the role of generative AI in transforming teaching, learning, and assessment practices. Journal of Applied Linguistics and TESOL. (JALT), 8(4), 159-174. https://doi.org/10.63878/jalt1314 Sun, Y. (2026). Beyond detection: GenAI in EAL writing education. ELT Journal, 80(1), 126–137. https://doi.org/10.1093/elt/ccaf054 Taylor, E. W. (2008). Transformative learning theory. New Directions for Adult and Continuing Education, 119, 5–15. Tran, P., & Dinneen, C. (2025). Reimagining EAP in the age of GenAI: innovation, integrity, and the future of assessment. University of Sydney Journal in TESOL, 4. https://doi.org/10.5281/zenodo.17590578 Vivas-Urias, M. D., Obispo-Díaz, C., & Ruiz-Rosillo, M. A. (2026). Analysis of teachers’ perceptions of the impact of Generative Artificial Intelligence in higher education. Journal of New Approaches in Educational Research, 15(1), 12. https://doi.org/10.1007/s44322-026-00060-5 Watson, C. E., & Rainie, L. (2025). Leading through disruption: higher education executives assess AI's impacts on teaching and learning. American Association of Colleges and Universities. http://files.eric.ed.gov/fulltext/ED671878.pdf Zlotnikova, I., Hlomani, H., Mokgetse, T., & Bagai, K. (2025). Establishing ethical standards for GenAI in university education: a roadmap for academic integrity and fairness. Journal of Information, Communication and Ethics in Society, 23(2), 188–216. https://doi.org/10.1108/JICES-07-2024-0104

Corresponding author: Karen K. Fujii Email: karenkf@hawaii.edu

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Appendix Semi-Structured Interview Questions Teaching Cohort 1. Please tell me about your role in the teaching cohort. 2. What views do your peers have about teaching or creating content with AI? How are your views similar or different to your peers? 3. What do you think influences how you are similar or different when it comes to AI compared with your teaching cohort peers? 4. What kinds of feelings or emotions have you experienced in working with this teaching cohort? 5. Have you noticed any shifts or things that have reinforced your teaching philosophy since working in this cohort? Teaching Experiences 1. What has your experience been like teaching students using AI-generated content? 2. Have your views or practices changed since you started working with AI-generated content? If so, how? 3. What has been the most surprising part of this experience for you? 4. If you were advising another instructor beginning to grade AI-generated work, what would you tell them based on your experience? 5. What kinds of feelings or emotions have you experienced in using AI to create content for your teaching? Grading Experiences 1. What has your experience been like grading students' work that are using AIgenerated content? a. What stood out to you in that experience? 2. How did that experience compare to grading student work that is not AI-assisted? 3. How do you decide what counts as acceptable or high-quality work when AI is involved?

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Exploring Hidden Structure and Improvised Practices Shaping Student Collaboration and Agency in Online Learning Farha Alia Mokhtar Universiti Malaysia Terengganu, Malaysia Sally Barnes University of Bristol, United Kingdom

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Abstract The rapid adoption of digital technologies in higher education has reshaped how students learn, collaborate, and engage, yet assumptions that online learning automatically enhances educational experiences remain largely unexamined. This study examined how students organised academic activities and appropriate digital tools within a fully online learning environment. Framed through a sociocultural perspective on mediated action, the study employed a qualitative case study design involving over 60 students and 30 hours of semistructured in-depth interviews with ten university students. Findings revealed an underexplored pattern of structured but non-linear group work, in which students improvised coordination practices through WhatsApp, Padlet, voice notes, calls, and peer reference. First-come, firstserved task allocation, volunteer compilation, and peer-based checking enabled students to organise group work but also contributed to uneven participation and limited opportunities for negotiation. Students further adapted their learning activities to home commitments, connectivity problems, and asynchronous communication, often prioritising task completion and submission over deeper content scrutiny. These findings demonstrate that student agency in online learning involves substantial coordination work, as learners construct and adapt social structures around available digital tools. However, such agency does not necessarily produce equitable or meaningful participation. The study contributes to sociocultural understandings of online learning by showing how students actively construct mediated collaborative practices when physical co-presence is absent, highlighting the need for pedagogical structures that support coordination, participation, and meaningful feedback. Keywords: digital affordances, higher education pedagogy, online learning, sociocultural theory, student agency

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

The adoption of digital technologies in higher education has increasingly transformed teaching and learning settings. These technologies offer various affordances that supposedly enhance student engagement, collaboration, and access to resources, thereby reshaping traditional educational paradigms. For example, the integration of hybrid learning models has allowed for flexible course delivery, afforded diverse student needs and promoted active learning (Ahmad et al., 2023). It must be noted that students’ interactions with the digital platforms used for their education were crucial, as these platforms served as mediational means through which students organised learning activities, communicated with peers, and experienced online learning. A case in point is Handel et al. (2020) who found that university students who were better equipped with technology, had more digital experience, and possessed higher self-reported digital learning skills experienced less tension, worry, and overload. Additionally, students who could communicate via digital devices reported lower levels of social loneliness. Similarly, Wong (2020) revealed that online learning encouraged students to develop autonomy and a sense of competence, enabling them to complete tasks independently. However, the effective inclusion of digital technologies requires not only the presence of technological tools but also strategic curriculum planning, professional development for teachers, and institutional evaluation of digital learning settings (Mokhtar et al., 2025). In the absence of intentional planning and supportive frameworks, the uptake of these tools might not reach their full capacity to improve educational results. For instance, the lack of social and physical interaction with peers and teachers led to solitary learning experiences, reduced motivation, and difficulty maintaining attention (Handel et al., 2020; Wong, 2020). The key issue is not merely the imposition of certain digital tools but rather how these tools can effectively support interactions and complement face-to-face learning both inside and outside the classroom. This implies that the shift towards digitalisation also carries its own set of challenges, such as ensuring the quality of online interactions and maintaining academic standards in virtual environments, despite offering its own set of affordances when learning takes place online. Consequently, this article problematises the assumption that digital technologies, particularly online learning, inherently improve educational experiences and explores the affordances and constraints of digital technologies in students’ online interactions, analysing the underlying processes of online learning and its influence on students’ education. Specifically, the research questions are as follows: 1. How do university students interact with digital tools to organise and complete academic tasks in a fully online learning environment? 2. What affordances and constraints do students experience when using digital technologies to support collaboration and learning in online settings? Literature Review Mediation in Teaching and Learning In sociocultural theory, mediation is the foundational concept that asserts humans do not interact with their environment directly; instead, their actions are shaped, transformed, and directed

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

through “mediational means” consisting of physical and psychological tools. Mediation describes human actions that employ these tools and language, which simultaneously shape the actions themselves (Vygotsky, 1981). Within this framework for education purposes, Lantolf and Thorne (2006) highlight that everyday interactions with these mediating artefacts are ubiquitous, offering students an indirect link to receive critical support (such as feedback, hints, and demonstrations) to complete tasks. This collaborative mediation typically operates as an interactional process between an expert (teacher or a more knowledgeable other) and a learner to guide them toward self-regulated activity, using situated feedback to promote development. Hence, in this study, mediation serves as the central concept used to understand and analyse how students and teachers interact with each other and their environment through physical and psychological tools to complete educational tasks (Wertsch, 1995). Affordances and Limitations of Digital Learning Digital tools have transformed educational environments by offering students greater flexibility and accessibility in learning. Online learning allows students to access resources and collaborate beyond traditional classroom settings, fostering participation across geographical and temporal boundaries. Digital platforms facilitate communication and information sharing, making it easier for students to engage in academic activities at their own pace. For instance, online learning can support students who may face physical or geographical barriers to education, while digital tools enable access to a wider range of educational resources (Lund & Rasmussen, 2008). These technologies can also provide opportunities for students who may be less vocal in face-to-face settings to participate and develop a sense of voice, belonging, motivation, as well as engagement within online learning communities (Adalberon & Säljö, 2017; Major et al., 2018). At the same time, digital technologies present constraints alongside their affordances, making an understanding of their capabilities and limitations important for designing effective technology-mediated learning experiences (Bond et al., 2021; Martin et al., 2020). The use of digital tools also enables students to take greater agency of their learning. Smørdal et al. (2021) and Rodness et al. (2021) observed that technology-mediated learning facilitated self-directed learning by allowing students to manage their learning processes based on their individual needs and preferences. However, the extent to which digital environments support such autonomy may also depend on how these platforms are designed and used. Research has suggested that virtual learning environments (VLE) are often used primarily as repositories for course materials rather than as interactive spaces for dialogue and collaboration (Crook & Cluley, 2009; McAvinia, 2016). These findings reveal a central tension in the literature on technology-mediated learning. Thus, digital technologies may increase learner autonomy and flexibility while simultaneously requiring students to take greater responsibility for organising their learning. A further gap concerns the distinction between the availability of digital tools and their actual appropriation in practice. Much of the literature evaluates whether particular technologies can facilitate engagement or learning, but less attention is given to how students negotiate and

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

repurpose these tools in naturally occurring educational settings. A tool’s intended pedagogical functions may differ from the way learners actually use it, because students may combine or adapt technologies in ways that were not anticipated by instructors or platform designers (Prieto et al., 2019). The resulting learning activity is therefore shaped not by technology alone, but by the interaction between human agency, social relationships, institutional expectations, and the cultural tools available to students (Prieto et al., 2019). From a sociocultural perspective, learner autonomy should therefore not be understood as freedom from others, but as a form of agency that develops through mediated participation with people, artefacts, and culturally organised practices. This perspective draws attention to an unresolved issue in the literature: the gap between the intended pedagogical affordances of institutional technologies and the ways students actually construct their own digital learning practices when existing platforms do not fully meet their collaborative needs (Lund & Rasmussen, 2008; Rodness et al., 2021; Smørdal et al., 2021). Accordingly, the present study examines technology use within a naturalistic, fully online learning environment, focusing on how students navigate the affordances and constraints of digital tools, coordinate their activities, and collectively construct learning practices. By shifting attention from technology as a predetermined intervention to technology as a resource actively appropriated by learners, the study responds to calls for greater understanding of how digital tools function within everyday educational environments (Bond et al., 2021). Constraints in the Mediated Nature of Online Learning Despite the affordances of digital tools, online learning presents challenges that can affect student engagement and learning outcomes. The absence of physical presence can create difficulties in maintaining interaction and fostering a sense of community among students. While digital platforms facilitate communication, they may not fully replicate the social and emotional connections that occur in face-to-face interactions. Research has shown that students in online learning environments may experience feelings of isolation and disengagement due to the lack of physical interaction with peers and instructors (Rasmussen et al., 2012). Evidence also suggests that participation varies according to learner characteristics and the conditions under which engagement is encouraged (Du et al., 2022; Goh et al., 2016). These findings suggest that the availability of digital spaces does not automatically lead to meaningful participation. Additionally, digital learning requires students to be self-motivated and disciplined, as the flexibility of online learning can sometimes lead to procrastination and reduced engagement. Another challenge is the credibility and reliability of information available through digital platforms. The abundance of online resources means that students must develop critical digital literacy skills to evaluate the accuracy and relevance of information. Without adequate guidance, students may struggle to distinguish between credible and unreliable sources, potentially affecting the quality of their learning (Lund & Rasmussen, 2008). Students may also experience uncertainty when evaluating the credibility of online information because webbased resources often come from diverse sources without clear indicators of authority (Crook,

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

2012; Major et al., 2018). This can contribute to frustration when learners encounter large volumes of information without sufficient scaffolding to guide their selection and interpretation (Knight & Mercer, 2015). Furthermore, the ease of accessing information online can sometimes lead to superficial learning practices, where students rely on readily available content without critically engaging with the material. This highlights the importance of digital literacy and appropriate scaffolding in helping students navigate online resources effectively. These findings point to a broader unresolved debate concerning the meaning of digital competence in higher education. Being able to access and operate digital tools does not necessarily imply that students are able to use those tools effectively for collaborative knowledge construction. Digital literacy is often conceptualised as an individual capacity, yet technology-mediated learning is inherently relational. Students must not only know how to operate a platform but also negotiate responsibilities, evaluate information collectively, coordinate contributions, and decide how digital resources should be used to achieve a shared academic purpose. This distinction is particularly important because students may demonstrate technological familiarity in everyday or social contexts while still experiencing difficulties in applying digital tools for critical thinking, academic collaboration, and purposeful learning (Knight & Mercer, 2015; Morrison, 2024). Thus, the central issue may not be whether students possess sufficient technological skills, but whether they can develop appropriate practices for using technology as a mediational means. This distinction remains understudied, particularly in naturally occurring learning environments. At the same time, instructors face related challenges in creating effective questioning strategies, providing appropriate resources, fostering collaboration, and managing students’ sense of isolation in online environments (Morrison, 2024). The integration of technical and pedagogical competencies can also be challenging for instructors as they work to establish new learning cultures in digital settings (Major et al., 2018). Additionally, the lack of face-to-face interaction can make it difficult for instructors to monitor student engagement and provide immediate feedback (Rasmussen et al., 2012). A related tension emerges between the individual and spaces in digital learning. Online environments are often valued for enabling learners to work independently and at their own pace, yet collaborative tasks require students to share tools, distribute responsibilities, negotiate meanings, and coordinate their actions across different temporal and social spaces. From a sociocultural perspective, this suggests that autonomy and collaboration are not opposing conditions. Rather, individual agency develops through participation in socially mediated activity, where learners construct and negotiate the practices through which digital tools become meaningful for learning. Pedagogical Structures in Online Versus Face-to-Face Learning The pedagogical structures in online learning differ significantly from those in face-to-face environments. In traditional classrooms, instructors play a central role in structuring learning activities, providing immediate feedback, and facilitating interactions among students. Faceto-face settings are typically characterised by synchronous interaction, where the simultaneous presence of instructors and students enables immediate exchanges, spontaneous discussion, and

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

access to non-verbal cues that can support learning (Lewohl, 2023). In contrast, online learning requires students to take more responsibility for managing their learning processes and engaging with digital tools. This shift in responsibility can be beneficial for developing students’ self-regulation and autonomy, but it can also create challenges for those who require more structured guidance and support (Rasmussen et al., 2012). The structural design of online learning therefore requires deliberate organisation of content, activities, and interaction, with digital tools potentially enhancing or constraining learning depending on their design and implementation (Liu et al., 2024). In face-to-face learning environments, social interactions occur naturally through physical proximity, allowing students to engage in spontaneous discussions and collaborative activities. In contrast, online learning environments rely on digital tools to facilitate interactions, which can sometimes result in delayed communication and reduced opportunities for spontaneous engagement (Lund & Rasmussen, 2008). However, the literature presents a somewhat simplified distinction between face-to-face and online learning by often treating physical presence and digital mediation as opposing conditions. Such a distinction risks overlooking the ways in which learners actively reconstruct social and pedagogical structures when physical co-presence is removed. The absence of face-to-face interaction may not eliminate social mediation; rather, it could redistribute the responsibility for creating and maintaining that mediation among students, teachers, and digital tools. This is particularly important because online learning requires more intentional strategies to foster participation and community when immediate interpersonal cues are unavailable. What is therefore less clear is how students themselves participate in reconstructing the social organisation of learning when established classroom structures are no longer available. This question also exposes a theoretical gap in how digital learning is commonly conceptualised. Research on educational technology frequently examines the effectiveness of individual platforms, tools, or instructional strategies, whereas sociocultural approaches invite attention to the activity system as a whole. From this perspective, a digital tool cannot be understood independently of the people who use it, the task they are attempting to accomplish, the institutional expectations surrounding that task, and the social relationships through which meaning is negotiated. The effectiveness of digital tools is therefore not necessarily determined by their technological features alone, as their educational value depends on how users perceive, adapt, and apply them within particular learning activities (Norman, 1988). Similarly, a VLE may be designed as a central learning platform but may become peripheral if students appropriate other tools that better support their immediate collaborative needs. The unresolved issue, therefore, is not simply which technology is most effective, but how people transform available tools into meaningful mediational resources for teaching and learning activities. The differences between online and face-to-face learning environments highlight the importance of pedagogical structures in shaping student engagement and learning outcomes. Examining these differences through the lens of mediated action may therefore provide a more nuanced understanding of how learning structures are produced and reproduced across different environments. Rather than assuming that online learning simply removes the structures

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

available in face-to-face classrooms, it is important to examine how students and teachers negotiate new forms of organisation through digital tools. In summary, the literature demonstrates that digital technologies offer greater flexibility, accessibility, and opportunities for collaboration, but their effectiveness depends on how they are integrated into pedagogy and appropriated by learners. Several gaps remain, particularly the tension between technology-enabled autonomy and the need for structure and coordination, the limited understanding of how students naturally adapt and repurpose digital tools, and the tendency to view digital competence as an individual rather than collective process. Accordingly, this study shifts attention from whether technology improves learning to how students collectively organise and appropriate digital tools in everyday online learning. By examining these processes, the study contributes to understanding how human agency and collaborative practices are shaped when learning is mediated through digital technologies in the absence of physical co-presence. Methodology Research Design A qualitative case study design was selected to facilitate an in-depth examination of subjective knowledge within the realm of education and learning (Crotty, 1998; Merriam, 2001). Rather than evaluating course content or learning outcomes, this study examined verbal and written interactions occurring within the course setting across face-to-face and fully online modalities. Although observations were conducted in both contexts, the analytical focus was the fully online learning environment, while face-to-face observations provided contextual and comparative evidence. Accordingly, aspects of sociocultural theory were integrated in this study to understand interactions in learning activities in the fully online learning context. This perspective highlights the importance of social and cultural dimensions while acknowledging individual agency within specific contexts (Wertsch, 1991). Sociocultural theory through mediation as a concept affords lens on group tasks as being a genuinely collective enterprise that take into account interplays between teacher, learners and cultural tools and that cannot be separated (except analytically) from learning activities. By focusing on interactions, the study examines how social, historical, and cultural aspects, along with the appropriation of tools, shape human learning experiences. Therefore, this study assumes that learning is part of a broader process of human transformation, where reality is socially constructed through ongoing interpretation of past events (Vygotsky, 1981). The agents involved students and instructors who utilise various cultural artefacts, with language being the primary mediating tool. Focusing on mediated activity avoids an overly individualistic approach and instead situates learning within institutional and historical contexts (Wertsch, 1995). Since descriptions alone do not fully reveal human-tool interactions, the analysis prioritises dynamic, explanatory insights into how mediation and affordances shape learning in digital spaces.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Research Context Data collection took place over approximately six months at a public tertiary institution in Malaysia. The same participants (described below) were observed in face-to-face classrooms and fully online learning contexts. Two voluntary instructors granted access to their classrooms comprised of university students in their second or final year who were enrolled in a mandatory English language course. This course aimed to develop students’ ability to apply language in social and professional contexts through tasks such as writing cover letters, emails, proposals, and participating in mock interviews. The university’s VLE was a mandatory tool for the course, while WhatsApp, Google Classroom, and Padlet were online tools and applications voluntarily chosen by the respective instructors, who deemed them suitable for their classes. It is pertinent to note that this research’s context took place during the global pandemic COVID-19 which reshaped the nature of mediation in teaching and learning activities by introducing a complex, multi-layered digital interface. Before the lockdown, classroom mediation occurred in a shared physical space where immediate verbal exchanges were supported by non-verbal interactional cues such as eye contact, nods, and physical gestures. During the fully online lockdown phase, this physical presence was lost, meaning that interactions had to be mediated entirely through digital devices and messaging platforms. The additional layer of technology-mediated communication filtered out vital physical cues, forcing students to actively adapt and structurally organise their virtual social spaces to successfully complete their coursework. More importantly, the study does not focus on the impacts of COVID-19; instead, it focuses on the environment within online educational spaces. It captures a unique opportunity to observe teaching and learning when it occurs fully online, as opposed to artificially requiring students to complete tasks online when they can still meet face-to-face. Participants The participants included two instructors, Awan and Bayu, along with two classes, each consisting of 30 to 36 final-year undergraduates. Both instructors consented to be observed and interviewed, while all student participants agreed to face-to-face and online learning observations, as well as audio recordings (Merriam & Tisdell, 2016). Additionally, convenience sampling was employed based on willingness to participate and availability, resulting in ten student participants for interviews and closer observation during task completion. This approach was considered suitable for capturing learning activities within a specific timeframe and context, particularly in examining interactions within the distinctive face-to-face and fully online learning environments. Data Sources The study sought to understand how and why particular interactional features of teaching and learning shaped students’ experiences across contrasting face-to-face and fully online environments. Data were collected through observations and semi-structured interviews.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

A non-participatory observation was adopted to maintain the naturalistic settings. This way, observations captured how learning activities unfolded, who participated, what tools were used, and when and where particular interactions occurred. Face-to-face classroom observations provided contextual evidence of how students and instructors organised tasks, communicated, and used available mediational resources in physical settings. In the online environment, observations included asynchronous interactions on Google Classroom, Padlet, and WhatsApp, as well as synchronous virtual learning sessions. The observations were important for identifying naturally occurring patterns of interaction and tool use that participants might not necessarily recall during interviews. At the same time, the researchers acknowledged that some aspects of online interaction remained less visible than face-to-face activity. Thus, the two observational contexts were connected analytically rather than treated as directly equivalent sources of data. Face-to-face observations provided a contextual and interactional reference point, while online observations constituted the primary empirical focus for examining digitally mediated learning activities. Semi-structured interviews were used to complement the observational data by addressing the how and why underlying participants’ actions and experiences. While observations could reveal how students interacted with teachers, peers, tasks, and mediational tools, interviews encouraged participants to reflect on their learning experiences, sources of support, and the ways in which particular tools or interactions shaped their engagement (Barlow, 2012). This was particularly important within a mediated action framework, as the meaning and effectiveness of a tool cannot be understood solely from its observable use. The study drew primarily on 14 semi-structured interviews with 10 students, totalling approximately 30 hours of individual interviews. Interview questions were informed (where relevant) by observed activities and interactions, allowing participants to clarify the intentions, meanings, and challenges behind actions that could be observed but not fully interpreted from the interactional record alone. All interview transcripts were provided to participants for member checking to verify the accuracy of both the native language versions and the English translations. Additional sources, including course documents and a researcher reflective journal, were used to contextualise the primary data, cross-check interpretations, and support reflexivity. Data Analysis Although data were collected across face-to-face and fully online settings, they were organised by data source and interactional context. Face-to-face observations provided contextual and comparative evidence, while online observations and interviews formed the primary analytical corpus. This distinction ensured that the face-to-face data were used to contextualise the findings rather than treated as an equivalent empirical focus. This enabled the researcher to trace how interactional practices were maintained, adapted, or transformed in the fully online environment. To analyse the data on the fully online learning modality, the study drew inspiration from thematic analysis approaches (Braun et al., 2019). Thematic analysis was used iteratively to examine student interactions in the online learning context. By coding language through key terms reflecting mediated activities, the study identified patterns in tool usage and learning behaviours.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Audio recordings from interviews and observation field notes were reviewed to identify emerging patterns, with transcriptions manually done for accuracy. Although online observations were rather limited because private student interactions were not always visible, but available threads on platforms like Google Classroom, Padlet, and WhatsApp provided insights into task execution. Data coding was conducted using thematic analysis, with initial manual coding subsequently managed through NVivo 11 for systematic data organisation and analysis. Interview transcripts and relevant observational data were coded into nodes such as priority, agency, and commitments, allowing coded extracts to be retrieved, compared, and refined across participants and data sources. NVivo also supported the iterative consolidation of codes into broader themes, while the final thematic relationships were interpreted in relation to the study’s mediated learning framework. Therefore, NVivo was used primarily for data management, coding, retrieval, and thematic comparison. Ethical Consideration and Positionality Ethical clearance for this study was obtained prior to data collection to ensure that potential physical and psychological risks were appropriately mitigated. Participants received a detailed research information sheet and provided written informed consent before data collection. Consent was consistently observed in the six-months period to ensure continued voluntary participation. Participants referenced in the dataset were assigned culturally and locally appropriate pseudonyms. Electronic data were securely stored using encrypted and passwordprotected systems, with identifiable data scheduled for permanent destruction following completion of the study. The researchers’ positionality was also considered to foster a fair and positive research environment. The researchers actively deemphasized institutional authority during fieldwork. Strategies included adopting culturally normative, informal familial forms of address, not wearing formal professional attire during fieldwork, providing refreshments during face-to-face interactions, and explicitly clarifying that research participation held no bearing on academic evaluation. These measures were intentionally deployed to establish rapport, built interpersonal trust, and create a psychologically safe space for participants, particularly students, who might otherwise hesitate to share authentic perspectives if they perceive an unsafe or authoritative environment. Findings Through a sociocultural lens, this study examined mediated interactions, uncovering patterns of task delegation, collaboration, and self-directed learning in digital spaces. The following section presents the key theme of Structured Social Space and its findings, highlighting the structured nature of online learning and how students navigated its affordances and constraints in their learning activities. Focusing on the fully online learning environments, this study finds ‘Structured Social Space’ as an umbrella theme that informs students’ processes to carry out group assignments, which were mainly executed through the WhatsApp virtual platform. Student participants’ use of available devices and features suggests appropriation of tool affordances and an enacted agency

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

to act with mediational means in performing their educational activities. Although much of the learning process was potentially more complex and covert than portrayed, the following subthemes highlight the visible interactions though which students worked together to ensure task completion in a fully online learning context. Students’ recollections were linked to two main topics taught in this fully online learning modality: a mock job interview and proposal writing. We drew on these topics as the focus of discussion to make sense of students’ educational activities. First-Come, First-Served Task Delegation and Collaboration Awan, the instructor in Class A, assigned a group project via WhatsApp. Upon receiving the task, Ramli started by spotting and adding group members into a newly created group chat before sending prompts to begin the assignment. According to Ramli, his initial prompt was phrased like an invitation. “Once we’ve identified our group members, we need to set up a WhatsApp group chat. Then, for example, I might say, “Hello, let’s begin — when are we starting on the assignment?” The grouping of students into the same chat upon identifying group members suggested that students were assigned by the instructor rather than self-selected. Ramli then initiated interaction with a greeting that called for a response, serving as an icebreaker before students could proceed to discuss the assignment. He followed up with “let’s go”, which acted as a prompt to his groupmates to start working. This message established a social context in the virtual space that encouraged participation, as the group’s progress on the assignment depended on members’ engagement. Suri observed that participation, such as suggesting topics for the assignment, came from the same few group members, while others simply agreed when asked. This highlights an imbalance in participation, with some students taking the lead while others remained passive. Usually, it’s the same few group members who suggest topics for the assignment. After that, we have to check with the others… most of them are just waiting around. When asked something like “is this okay?” they’ll respond, “Sure, I’ll just go along with it.” (Suri) The above extract illustrated how some students adopted a “just follow” approach, with limited participation in decision-making. The absence of a shared physical space may have shaped how students participated, contributing to more passive forms of engagement. Two common methods emerged for task delegation. Didi explained that group members would list assignment tasks in point form within their WhatsApp chat, leaving brackets for students to self-assign their preferred tasks. Meanwhile, Maya shared that since all members were already aware of the assignment requirements, they would voluntarily choose their parts and post their selections in the chat. Latecomers, she noted, would take on whichever tasks remained.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Either I or someone else will list the main points that need to be included in the group assignment and place brackets next to each one. So, for instance, if someone wants to take on task A, they’ll just write their name in the bracket. Once tasks like point 1 and 2 are taken, the remaining points go to the rest of the group. (Didi) When we split the work, we usually already know all the components that need to be covered, so we just volunteer — like, who wants to handle what. Anyone who joins late will take on whatever is left. That way, everyone ends up with their own part. (Maya) The delegation methods highlighted a structured approach to virtual collaboration, where tasks were listed and selections were posted to ensure clarity on available roles and remaining gaps. These processes suggested that some students worked individually on assignment requirements before coordinating as a group in the chat. However, the first-come, first-served task delegation structure disadvantaged students who lacked the opportunity to negotiate their roles. For instance, Orked suspected that certain teammates had already pre-assigned tasks among themselves before formally announcing them in the chat, limiting her choices and potentially affecting group dynamics. It seems like they’ve already made decisions among themselves — like one person will do this section, another will take that part. So by the time I checked the messages, they had already assigned the parts, and I had to just pick from what was left. (Orked) The extracts above highlighted constraints in the virtual delegation system. While allowing students to select tasks independently was efficient, as it enabled them to work separately, it also led to dissatisfaction due to limited negotiation, lack of transparency, and inadequate communication about assignment expectations. The first-come, first-served task delegation and collaborative approach also disadvantaged students with home commitments. These students might have missed the incoming messages because there was an assumption that they always had access to their phones and the internet and were expected to be constantly engaged. The analysis in this sub-theme revealed two key sociocultural aspects of virtual delegation. First, students’ agency played a crucial role, as active members initiated discussions and structured the workflow. Second, the necessity for a structured virtual space emerged, ensuring all members stayed informed about task progress and assignment expectations. Evidently, the virtual setting required explicit communication to clarify roles and responsibilities. This structure offered both advantages and disadvantages, as it helped students navigate group work in isolation, thereby increasing awareness of their contributions and responsibilities.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Fusion of Physical and Virtual Spaces in Student Engagement Putri illustrated her experience working on a Mock Job Interview group assignment assigned by instructor Bayu in Class B. She recalled having to work at night, as it was the only time all group members were available to work on the assignment. This assessment involved groups of three taking turns as interviewers and interviewees. In this fully online learning modality, the assignment was adapted into a video recording, which had to be submitted on Padlet (a collaborative education website) by the given deadline. For this mock interview assignment, we scheduled the morning for everyone to go over videos and notes on Padlet… and we saved the night for recording. Most of the time, we could only find a slot at midnight when everyone was finally available. (Putri) As seen in the above extract, Putri described how her group coordinated their schedule by allocating the morning to reviewing shared videos and materials on Padlet, while reserving the nighttime for completing the group task. This arrangement reflected how students balanced home commitments with academic responsibilities, emphasizing the importance of time management and the strategic use of available resources. Similarly, Nurin from Class A faced a challenge when she found herself unprepared for her scheduled Mock Job Interview assessment, which was to take place on Google Meet with the instructor, Awan, as the interviewer. To address this, she reached out to her peers, Datul and Fasha, to organise a last-minute practice session. Datul took the initiative in coordinating the discussion, suggesting a suitable time and platform. Before meeting, all members searched for interview tips to ensure they were well-prepared. The practice session was conducted on the evening of the announcement, allowing them to refine their responses and boost their confidence before the assessment. We were assigned a mock interview unexpectedly — our instructor gave the task at night, and the assessments were already happening the next day… but we didn’t get to rehearse much. Datul said, it’s okay, let’s just talk about it here. We all searched for tips on answering interview questions, then we practiced that same night. (Nurin) Nurin’s account in the extract above described how students adapted to home-based learning by prioritising household commitments and allocating specific periods at night for academic tasks. In both cases, students demonstrated commitment and peer support through selforganised study groups, urgent decision-making, role delegation, and rapid information gathering within a limited timeframe. Nurin’s experience also highlighted constraints on students’ ability to negotiate schedules and seek clarification from instructors in real time. Given the absence of face-to-face interactions, students utilised other digital affordances on their devices, such as video calls. However, video calls were not without their own challenges during learning activities. Orked recalled Suri struggling with connectivity issues during the Mock Job Interview recording, causing her screen to lag and become unresponsive. Because

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

rerecording was too time-consuming, the group opted to use the available recording despite an unavoidable blooper at the end. While we were recording, suddenly Suri’s mouth was open and she went silent, so Putri and I tried not to laugh… we didn’t edit the video or cut anything using a laptop because it was too much hassle… so the final part got included, showing Suri with her mouth wide open, looking funny. (Orked) The extract above highlighted how students adapted to fully online learning despite technological limitations. While video calls allowed re-recording, constraints such as scheduling, internet reliability, video editing, and submission deadlines led students to use available materials and prioritise timely submission. Other affordances used to cope with the physical context’s demands were voice notes and voice calls. While voice notes supported learning, they also led to issues like missed conversations and poor sound quality. To address this, students used voice calls for clearer communication. Didi noted that calls helped gather everyone, facilitate detailed discussions, and clarify matters effectively. When there are too many voice notes — let’s say we’ve had a long discussion on WhatsApp, and one member didn’t notice the voice messages because they weren’t online… it becomes difficult for them to catch up. Some voice notes are also hard to hear. But when we do a live call, it becomes a space where everyone can come together. If we need to ask something, we ask directly and write down what we’ve talked about. (Didi) On WhatsApp, if someone has a bad signal, and we’ve already had a group discussion, then they log in much later, it’s a pain for them to scroll all the way up. What if they miss something important that they need to understand before they can do their task? (Ramli) The immediacy of feedback in discussions suggested that direct verbal exchanges were more effective for resolving ambiguities. It showed that students actively compensated for the loss of interactional immediacy by selecting tools that reproduced particular communicative functions. While calls connected people despite physical separation, students also highlighted the structural need for coordination, as members had to be invited or added before participating. The findings in this sub-theme highlighted two key aspects of student interactions. First, students engaged in both individual and group interactions. Individually, they accessed and processed digital resources, such as learning materials on Padlet and job interview tips, before presenting their findings in group discussions. This underscored the heightened reliance on digital tools in remote learning contexts. Second, student interactions were shaped by their home environments. For instance, working at night due to a more conducive setting reflects the interplay between physical surroundings and virtual learning. The interdependence between

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

these contexts suggested that while online learning provides access to education, it remains closely tied to students’ physical conditions, directly influencing their engagement and ability to participate effectively. Submission Priority Versus Content Comprehension The majority of students explained that at the end of their group work, one member would voluntarily compile the assignments and share the document for all members’ review. Members who received comments were expected to amend their work before resubmitting. Usually, one person will compile the assignment, combining everyone’s parts into one final document… then we check which parts have errors — sometimes things are copypasted incorrectly, or there are grammar issues, and so on. (Didi) We send our parts in the group chat, then one of our group mates puts it all together and shares the completed version back in the chat. Once everyone agrees that it’s fine, then the file gets submitted. Most of the time, we just compile the parts. (Lana) The interview extracts revealed a structured virtual workflow where a volunteer compiled members’ individual contributions. The compiled assignment was then shared for review, with all members expected to approve the final version before submission. This process suggested a sense of shared accountability, reinforced by visible feedback that added a layer of crosschecking. Nonetheless, students primarily focused on submission rather than content scrutiny, with checks limited to formatting and grammar. This was evident in Lana’s statement, “Most of the time we will only compile,” and Suri’s admission that she skimmed through the work without detailed review before submission. In contrast, Maya emphasized the importance of thorough checking, believing that collective evaluation requires a final outcome that is satisfactory to all. It’s a group assignment, so we don’t want anyone submitting a mediocre piece of work, because it affects everyone’s grade. (Maya) For individual assignment checks, students in Class B stated that Padlet, the designated platform for submitting formative assessments, enabled them to study their peers’ work before submission. Suri further explained that after reviewing her peers’ work on Padlet, she would reach out to Orked and Putri to share her observations and clarify necessary improvements for their assignments. I usually wait to see what others have submitted on Padlet, and I’ll check their assignments before I improve mine. (Putri) I scroll through Padlet to read my classmates’ submissions. If I see something in theirs that I didn’t include in mine, I’ll add it into my own work. (Zain)

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

On Padlet, I review other classmates’ assignments, then I’ll let Orked and Putri know what we need to add into ours — I’d say something like “okay, do it this way.” Then we’ll submit. I have to see how others do theirs first so I know how to approach mine. (Suri) Padlet’s vertical layout made it convenient for students to scroll down and check their peers’ work because the instructor had organised each column by topic (see Figure 1 below): Figure 1 Student Formative Assessment Submissions and Peer Review Activity on Padlet

The interview extracts indicated a pattern in which students used Padlet as a reference tool for assignment checking. In addition to being structured by its layout and design, the platform served as a repository of information, allowing students to compare and contrast their work with their peers’ submissions to improve their assignments before submission. Additionally, it gave students a clearer idea of the expected assignment outcome, reducing the challenge of producing work from scratch at home. However, relying on peers’ submissions for reference was not always feasible, as waiting too risked missing the one-day deadline due to preparation and task execution. The analysis revealed a gap between students’ ideal group work—checking content quality— and their actual practice, which focused on technical aspects like grammar and formatting. This indicated that students’ primary priority in the fully online learning context was task submission rather than deep engagement with the assignment. The structured workflow for group assignments relied on a volunteer compiler who gathered, shared, and submitted the final document, reinforcing submission-driven interactions. Meanwhile, individual assignments required students to independently review their peers’ work and identify inspiring examples to inform their own work.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

A visual summary of this covert structure within the mediated online learning context is presented in Figure 2 below: Figure 2 Process Map of Structured but Non-Linear Group Work in Online Learning

The process map above shows how university students navigate group assignments in a fully online learning environment. Instead of following a linear progression, their workflow was moulded by both digital affordances and contextual constraints, resulting in an iterative and dynamic process. It begins with group formation and task prompts via WhatsApp, followed by asynchronous idea sharing and topic suggestions where a few active members typically lead. Next is self-assigned task delegation, often on a first-come, first-served basis, which can limit negotiation and transparency. Content review and reference to peers’ work happen through platforms such as Padlet to guide improvements and alignment. At the same time, group discussions occur via voice notes or synchronous calls, particularly when clarity is needed. Final compilation and group review are carried out by a volunteer before submission, which emphasises the idea of submission priority over content comprehension. Throughout the execution of group work, students balance academic tasks with home responsibilities, often working at night and adapting creatively despite technological limitations. This cyclical and collaborative structure challenges the assumption that online learning is straightforward and instead shows how students construct shared learning spaces and coordinate responsibilities in complex, multimodal ways.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Discussion The findings revealed that fully online learning required students to develop more explicit structures for group activity. However, these structures were not linear or institutionally prescribed. Rather, students improvised organisational practices across WhatsApp, Padlet, voice notes, calls, individual tasks, and peer consultation. This pattern of structured but nonlinear group work shows how students constructed their own coordination processes in response to the demands and constraints of the online environment. In addressing the first research question, the study shows that students interacted with digital tools through taskoriented workflows involving group formation, role allocation, discussion, peer reference, compilation, and submission. Leadership roles also emerged more explicitly, as some students took the initiative to guide their peers in structuring assignments (Lund et al., 2019). The significance of this finding lies in showing that digital environments do not simply transfer existing learning practices into virtual spaces. Rather, the removal of physical co-presence appears to make aspects of collaborative learning that are often implicit in face-to-face settings more explicit and operationalised. Students therefore had to actively construct the organisational structures needed to sustain collaboration, demonstrating that learner agency becomes particularly important when interaction is mediated through digital tools. Both face-to-face and VLEs involved mediation in learning activities, yet the nature of mediation differed significantly (Vygotsky, 1981). In face-to-face settings, delegation and leadership took place in a shared physical space, where real-time verbal and non-verbal cues such as gestures, facial expressions, and eye contact, all facilitated nuanced communication and collective decision-making (Lewohl, 2023). These multimodal interactions contributed to more fluid learning experiences. In fully online learning environments, interactions were mediated through digital devices and messaging applications to recreate a sense of presence in the absence of face-to-face interactions. This shift highlights that students’ organisation of academic tasks relied heavily on explicit communication, visible coordination, and the externalisation of processes that would otherwise remain implicit in physical settings. For example, students supplemented university learning management systems with WhatsApp video calls to facilitate real-time communication. The platform played a central role, not merely as a communication tool but as a mediator for learning activities, demonstrating how students adapted non-university tools to meet their academic needs while simultaneously exercising autonomy in their use of digital platforms. This finding is important from a sociocultural standpoint because it shows that learning is not mediated exclusively by institutionally provided technologies. Students actively appropriate available tools according to the demands of the activity, suggesting that the effectiveness of a digital learning environment is partly shaped by students’ capacity to exercise agency and by the affordances of the broader technological ecosystem in which learning takes place. At the same time, this adaptation raises questions about whether institutional learning environments sufficiently accommodate the collaborative practices students naturally develop.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

In relation to the second research question, the situation above reflects a key affordance of digital technologies: namely, the flexibility and accessibility offered across multiple platforms that support continuous collaboration. However, these affordances were accompanied by significant constraints. Students found it difficult to gauge their peers’ immediate thoughts and intentions, as digital interactions could not fully capture body language. Text-based communication further reduced interactional cues, leading students to establish structured protocols to ensure a shared understanding of tasks (Adalberon & Säljö, 2017). The uptake of other digital tools beyond the university’s VLE emphasizes the VLE’s limitations in supporting students’ online education requirements. The findings, therefore, suggest that technological access alone does not guarantee meaningful interaction. Rather, students’ experiences were shaped by a tension between the flexibility offered by digital technologies and the interactional limitations introduced by reduced social and non-verbal cues. This is important to highlight because it challenges assumptions that online learning is inherently more accessible or flexible simply because learning materials and communication can be accessed across time and space. Flexibility may facilitate participation, but without appropriate interactional support, it may also require students to undertake additional organisational and communicative work to sustain collaboration. The asynchronous nature of online learning also shaped this structured but non-linear pattern of interaction. Without immediate time constraints, students gained flexibility to allocate time for their tasks and extend discussions over longer periods. However, synchronous online meetings required students to coordinate availability and take turns speaking, as digital platforms could not accommodate overlapping voices without disruption (Crook, 2012). Consequently, digital interactions required clearer organisation, making task delegation and participation more visible. These findings indicate that digital mediation does not merely constrain or enable learning in isolation. Instead, it reshapes the conditions under which students organise, communicate, and exercise agency. The observed practices suggest that students developed compensatory strategies to address the limitations of digital interaction, but these strategies also placed greater responsibility on learners to coordinate their own participation and sustain group processes. The broader implication is that sustainable integration of technology in education cannot rely solely on student adaptation (Liu et al., 2024). Despite these noteworthy findings, several limitations should be acknowledged. First, the study was situated within a specific institutional and educational context, which means that the interactional patterns observed may not be directly transferable to other universities, disciplines, or student populations. Second, the analysis focused on selected digital platforms and observable interactions, meaning that some aspects of students’ online learning experiences may have remained inaccessible to the researcher, particularly private interactions or activities conducted through platforms outside the scope of observation. Third, while the study provides insight into how students organised and experienced digitally mediated learning, it does not establish whether particular interactional patterns directly resulted in improved learning outcomes. The findings should, therefore, be interpreted primarily as an account of how students navigated and negotiated learning within a digitally mediated

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

environment, rather than as evidence of the relative effectiveness of online or face-to-face learning. Conclusion Online learning is not simply a matter of transferring existing teaching and learning practices into a digital environment. The findings demonstrate that digital tools introduce an additional layer of mediation that reshapes how students communicate, coordinate tasks, participate, and engage with learning. Although platforms such as Google Classroom, Padlet, WhatsApp, and video-conferencing tools enabled students to organise tasks and maintain interaction across physical and virtual spaces, the observed interactions also revealed communication breakdowns, uneven participation, temporal mismatches, and forms of engagement that were sometimes oriented towards meeting submission requirements rather than developing deeper understanding. At the same time, students demonstrated agency through self-organisation, peer coordination, and independent problem-solving. These findings suggest that the educational value of online learning depends not simply on the presence of digital technologies, but on how students, teachers, tasks, and technological tools are socially and pedagogically organised. Accordingly, online environments should not be assumed to reproduce the interactional immediacy of face-to-face learning. Instead, digital technologies should be understood as mediational resources whose effectiveness depends on intentional pedagogical design, appropriate interactional structures, and opportunities for meaningful feedback. Based on these findings, online learning environments should incorporate more deliberate structures for collaborative participation. The observed reliance on first-come, first-served task allocation and uneven contribution suggests that educators could introduce clearly defined group roles, transparent task responsibilities, and structured opportunities for peer interaction. Communication breakdowns and limited opportunities for immediate clarification also indicate the need for more timely and accessible feedback through guided discussion prompts, scheduled check-ins, and targeted feedback points. To address the tendency towards task completion rather than deeper engagement, online activities could incorporate collaborative problem-solving, reflection, and opportunities for students to explain and justify their decisions. At the institutional level, professional development should equip instructors not only with technical skills but also with pedagogical strategies for facilitating interaction and feedback across synchronous and asynchronous environments. These recommendations are directly grounded in the interactional patterns observed in the study and should therefore be understood as practical considerations for educators working in comparable contexts rather than universal prescriptions. The study also highlights several directions for future research. As the research was situated within a single institutional context, future studies could examine multiple institutions and student populations to explore how sociocultural, disciplinary, and technological differences shape mediated learning experiences. Further research could also examine students’ broader digital ecosystems, as the present study focused on interactions visible through selected platforms and could not capture every aspect of their online learning experiences. In addition,

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

combining qualitative interactional evidence with quantitative measures of academic performance could provide stronger insight into the relationship between interaction patterns and learning outcomes. Longitudinal research could further examine whether students’ adaptive strategies contribute to sustained learning, autonomy, and digital agency over time. Overall, the findings indicate that the key question is not whether digital technologies should be used in higher education, but how they should be intentionally designed and orchestrated. Students can adapt creatively to digitally mediated environments, but their agency alone cannot compensate for weak interactional structures, limited feedback, or poorly designed activities. Meaningful online learning therefore requires deliberate alignment among learners, educators, tasks, and digital tools.

Acknowledgements This work was supported by the SLAM scheme, Universiti Malaysia Terengganu, Malaysia. We thank the participants for their contribution in this research. Declaration of Generative AI and AI-assisted technologies in the writing process The authors would like to acknowledge the use of artificial intelligence (AI) tools in the preparation of this manuscript. The following tool was used to enhance language and readability purposes only: ChatGPT. All AI-edited text was thoroughly reviewed and revised by the authors to ensure accuracy, clarity, and adherence. AI tools were not used for other purposes, including data generation, data analysis, methodology, or interpretation of findings or conclusions. The authors take full responsibility for all aspects of the final manuscript.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

References Adalberon, E., & Säljö, R. (2017). Informal use of social media in higher education: A case study of Facebook groups. Nordic Journal of Digital Literacy, 12(4), 114–128. https://doi.org/10.18261/issn.1891-943x-2017-04-02 Ahmad, S., Umirzakova, S., Mujtaba, G., Amin, M. S., & Whangbo, T. (2023). Education 5.0: Requirements, enabling technologies, and future directions [Preprint]. arXiv. https://doi.org/10.48550/arXiv.2307.15846 Barlow, C. A. (2012). Interviews. In Sage research methods: Encyclopedia of case study research (pp. 496–499). SAGE Publications. Bond, M., Marin, V. I., Dolch, C., Bedenlier, S., & Zawacki-Richter, O. (2021). Digital transformation in higher education: EdTech perspectives on technology-enhanced learning. Educational Technology Research and Development, 69(2), 1–24. https://doi.org/10.1007/s11423-021-09988-2 Braun, V., Clarke, V., Hayfield, N., & Terry, G. (2019). Thematic analysis. In P. Liamputtong (Ed.), Handbook of research methods in health social sciences. Springer. Crook, C. (2012). The ‘digital native’ in context: Tensions associated with importing Web 2.0 practices into the school setting. Oxford Review of Education, 38(1), 63–80. https://doi.org/10.1080/03054985.2011.577946 Crook, C., & Cluley, R. (2009). The teaching voice on the learning platform: Seeking classroom climates within a virtual learning environment. Learning, Media and Technology, 34(3), 199–213. https://doi.org/10.1080/17439880903141570 Crotty, M. (1998). The foundations of social research: Meaning and perspective in the research process. SAGE Publications. Du, Z., Wang, F., Wang, S., & Xiao, X. (2022). Enhancing learner participation in online discussion forums in Massive Open Online Courses: The role of mandatory participation. Frontiers in Psychology, 13, Article 819640. https://doi.org/10.3389/fpsyg.2022.819640 Goh, C. S., Tan, T. G., & Ahmad Buhari, T. (2016). Learners’ attitudes towards engaging in online communication activities. International Journal of e-Learning and Higher Education, 5(4), 43–54. Handel, M., Stephan, J., Gläser-Zikuda, M., Kopp, B., Bedenlier, S., & Ziegler, A. (2020). Digital readiness and its effects on higher education students’ socio-emotional perceptions in the context of the COVID-19 pandemic. Journal of Research on Technology in Education, 54(2), 267–280. https://doi.org/10.1080/15391523.2020.1846147 Knight, S., & Mercer, N. (2015). The role of exploratory talk in classroom search engine tasks. Technology, Pedagogy and Education, 24(3), 303–319. https://doi.org/10.1080/1475939X.2014.931884 Lantolf, J. P., & Thorne, S. L. (2006). Sociocultural theory and the genesis of second language development. Oxford University Press.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Lewohl, M. J. (2023). Exploring student perceptions and use of face-to-face classes, technology-enhanced active learning, and online resources. International Journal of Educational Technology in Higher Education, 20(1), 48. https://doi.org/10.1186/s41239-023-00416-3 Liu, Q., Chen, L., Feng, X., Bai, X., & Ma, Z. (2024). Supporting students and instructors in blended learning. In Handbook of educational reform through blended learning (pp. 45-60). Springer. https://doi.org/10.1007/978-981-99-6269-3_5 Lund, A., Furberg, A., & Gudmundsdottir, G. B. (2019). Expanding and embedding digital literacies: Transformative agency in education. Media and Communication, 7(2), 47– 58. https://doi.org/10.17645/mac.v7i2.1880 Lund, A., & Rasmussen, I. (2008). The right tool for the wrong task? Match and mismatch between first and second stimulus in double stimulation. International Journal of Computer-Supported Collaborative Learning, 3(4), 387–412. https://doi.org/10.1007/s11412-008-9050-8 Major, L., Warwick, P., Rasmussen, I., & Ludvigsen, S. (2018). Classroom dialogue and digital technologies: A scoping review. Education and Information Technologies, 23(5), 1995–2028. https://doi.org/10.1007/s10639-018-9701-y Martin, F., Sun, T., & Westine, C. D. (2020). A systematic review of research on online teaching and learning from 2009 to 2018. Computers & Education, 159, Article 104009. https://doi.org/10.1016/j.compedu.2020.104009 McAvinia, C. (2016). Online learning and its users: Lessons for higher education. Chandos Publishing. Merriam, S. B. (2001). Andragogy and self-directed learning: Pillars of adult learning theory. New Directions for Adult and Continuing Education, 2001(89), 3–13. https://doi.org/10.1002/ace.3 Merriam, S. B., & Tisdell, E. J. (2016). Qualitative research: A guide to design and implementation. Jossey-Bass. Morrison, R. (2024). Making the invisible visible: Critical discourse analysis as a tool for search engine research. Journal of the Association for Information Science and Technology, 75(5), 600–612. https://doi.org/10.1002/asi.24859 Mokhtar, F. A., Arshad, S., Jamil, N. J., Zakaria, M. K., Wan Ibrahim, C. W. I. R. C., Shamsudin, C. M., & Awang, S. (2025). Teacher–student interactions in face-to-face and online learning: A sociocultural case study in Malaysian higher education. International Journal of Learning, Teaching and Educational Research, 24(12). https://doi.org/10.26803/ijlter.24.12.26 Norman, D. A. (1988). The psychology of everyday things. Basic Books. Prieto, L. P., Rodríguez-Triana, M. J., Martínez-Maldonado, R., Dimitriadis, Y., & Gašević, D. (2019). Orchestrating learning analytics (OrLA): Supporting inter-stakeholder communication about adoption of learning analytics at the classroom level. Australasian Journal of Educational Technology, 35(4). https://doi.org/10.14742/ajet.4314 Rasmussen, I., Lund, A., & Smørdal, O. (2012). Visualisation of trajectories of participation in a wiki: A basis for feedback and assessment? Nordic Journal of Digital Literacy, 7(1), 20–35. https://doi.org/10.18261/ISSN1891-943X-2012-01-03

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Rodness, K. A., Rasmussen, I., Omland, M., & Cook, V. (2021). Who has power? An investigation of how one teacher led her class towards understanding an academic concept through talking and microblogging. Teaching and Teacher Education, 98, Article 103229. https://psycnet.apa.org/doi/10.1016/j.tate.2020.103229 Smørdal, O., Rasmussen, I., & Major, L. (2021). Supporting classroom dialogue through developing the Talkwall microblogging tool: Considering emerging concepts that bridge theory, practice, and design. Nordic Journal of Digital Literacy, 16(2), 50-64. https://doi.org/10.18261/issn.1891-943x-2021-02-02 Vygotsky, L. S. (1981). The genesis of higher mental functions. In J. Wertsch (Ed.), The concept of activity in Soviet psychology. M.E. Sharpe. Wong, R. (2020). When no one can go to school: does online learning meet students’ basic learning needs? Interactive learning environments, 31(1), 1–17. https://doi.org/10.1080/10494820.2020.1789672 Wertsch, J. V. (1991). Voices of the Mind: Sociocultural Approach to Mediated Action. Harvard University Press. Wertsch, J. V. (1995). The need for action in sociocultural research. In J. V. Wertsch, P. del Rio, & A. Alvarez (Eds.), Sociocultural Studies of Mind (pp. 56–74). Cambridge University Press.

Corresponding author: Farha Alia Mokhtar Email: alia.mokhtar@umt.edu.my

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Prompting Intentionality and Authorship Preservation in AI-Assisted EFL Academic Writing: A Process-Tracing Inquiry Rym Ladjal Laboratoire de Linguistique et Sociodidactique du Plurilinguisme LISODIP. École Normale Supérieure de Bouzaréah- Chikh Moubarek BenMohammed Ibrahimi Elmili Eldjazairi, ENSB. Algiers, Algeria Hayat Messekher École Normale Supérieure de Bouzaréah - Chikh Moubarek Ben Mohammed Ibrahimi Elmili Eldjazairi, Algeria

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Abstract Large language models (LLMs) are reshaping how researchers write and negotiate authorship. As these tools grow increasingly capable of producing fluent academic prose, a pressing question emerges: When writers use LLMs to help produce an academic text, how much of their thinking survives the exchange? Research has examined attitudes toward generative artificial intelligence (GenAI) and the quality of AI-assisted outputs; however, the behavioural mechanisms through which writers maintain or relinquish authorial control remain underexplored. This study introduces prompting intentionality and examines its association with authorship preservation. Sixteen EFL doctoral researchers completed a five-stage processtracing protocol progressing from human-only ideation through naive, guided, and strategic prompting to final re-authoring. Data were collected from six sources, including prompt texts, conversation threads, authorship trace annotations, reflections, and a Likert-scale instrument, and analysed using a convergent mixed-methods design. Prompting intentionality rose across the three stages (mean 0.31 to 3.56; Friedman χ² = 25.20, p < .001) and was associated with authorship preservation (rs = .76, p < .001; mean Human Agency Preservation Ratio 79.4%). Rather than a uniform rise, criteria showed a clear ordering: Writers encoded personal stance and content readily but assigned the tool a bounded role least often. Strategic prompting was identified as the stage of strongest authorial control, though many did not revise AI output, indicating that agency was exercised more at encoding than at output selection. An emergent conceptual pattern, the Agentic Prompting Loop, is proposed to describe this negotiation of human agency. The findings underscore the importance of explicit prompting literacy instruction for responsible GenAI integration. Keywords: AI-assisted EFL academic writing, human agency, authorship preservation, epistemic authority, prompting intentionality, responsible AI integration

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Scholarly writing has historically been dialogic. Researchers engage with existing literature, identify conceptual gaps, and position their arguments within ongoing scholarly conversations. Although shaped by prior knowledge, this process remains interpretive: meaning is synthesized internally, and argumentative structure often emerges through cognitive negotiation (Flower & Hayes, 1981). The emergence of large language models (LLMs), however, may alter this process for some writers. Rather than moving sequentially from interpretive reading to authorial positioning and drafting, writers may begin with algorithmically generated summaries, outlines, or argumentative frames. In this sense, rhetorical organization may increasingly precede, or shape, internal conceptual formulation. This shift does not render generative artificial intelligence (GenAI) tools inherently problematic. However, it raises a foundational question: at what stage does authorship originate? If argumentative structures are externally generated before the writer articulates an independent conceptual stance, authorship may become partially displaced. This displacement concerns writers, who must be able to account for their own contribution; educators, who assess work of increasingly opaque provenance; and publishers, who require authors to take responsibility for content they may not have originated. This study examines the pre-structural authorship stage, the moment at which writers formulate an initial thesis or argumentative intention before GenAI interaction, and investigates how this stage is preserved, negotiated, or diminished during AI-assisted academic writing. In this study, this stage is operationalized as a mandatory human-only ideation phase completed before participants interact with any GenAI tool. Research on AI-assisted writing has expanded rapidly, yet its trajectory reveals an unresolved tension. Early studies focused primarily on perceptions, examining how students and educators view GenAI tools and their pedagogical value (Sanz-Tejeda et al., 2026). While necessary, these attitudinal studies left a more urgent question unanswered: when writers use GenAI to produce an academic text, how much of their intellectual DNA survives the interaction? The answer determines whether a writer can legitimately claim authorship of the resulting text, and whether they retain the epistemic authority that scholarly writing presumes. In this study, intellectual DNA refers to the combination of prior knowledge, developed arguments, disciplinary positioning, and epistemic judgment that marks a text as recognizably one’s own. The concern is not that GenAI produces effective writing. Rather, writers who engage without deliberate intentionality risk producing texts that are fluent yet intellectually hollow, structurally competent yet detached from their own thinking (Kim et al., 2025). This risk is amplified by AI detection tools: minimal prompting may yield outputs more likely to be flagged as AI-produced, not through intended deception but because the writer surrendered their intellectual DNA at the prompting stage. Consider two doctoral researchers completing the same writing task with the same LLM. The first submits a brief prompt and accepts the output with minor revisions. The second articulates

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

a personal position, determines which arguments require development, assigns the GenAI a limited role, requires verifiable sources, and instructs it to flag uncertainty rather than hallucinate. Same tool, same task, yet only one meaningfully remains the epistemic author. The second prompt demands more time and deliberation than the first, and that difference and additional investment are what this study measures. This distinction is conceptualized in the present study as prompting intentionality: the degree to which writers encode their reasoning, stance, and epistemic decisions into the prompt before the LLM responds. Prompting intentionality is not a technical skill but an authorial one, not a fixed attribute, but a developable and assessable behaviour (Zamfirescu-Pereira et al., 2023). Despite its relevance to authorship and agency, the relationship between what writers bring to the prompting act, what the LLM produces in response, and how much of the writer’s thinking survives in the final text remains under-examined. This study seeks to investigate that relationship to examine whether prompting intentionality can be assessed and taught, and whether it may function as a mechanism through which writers retain authorship of AI-assisted texts. This inquiry is particularly urgent in the present context. English as a foreign language (EFL) doctoral researchers face a dual challenge: as emerging scholars, they are expected to produce original research bearing their own epistemic authority, yet as EFL writers, they may rely on GenAI tools for language support. Without intentional prompting, however, these tools may displace the writer’s voice precisely where it carries the greatest scholarly value (Tour & Zadorozhnyy, 2025). The study has two objectives. Conceptually, it examines how prompting intentionality develops across three prompting conditions and how it is associated with authorship preservation. Practically, it develops a purpose-built rubric for assessing prompting intentionality, as no validated instrument currently exists. Using both behavioural and attitudinal evidence, the study addresses these objectives together. The study employs a staged process-tracing protocol with sixteen EFL doctoral researchers at the Teacher Training College (École Normale Supérieure de Bouzaréah, ENSB), a higher education institution in Algiers, Algeria. Three research questions guide the inquiry. First, how does prompting intentionality develop across naive, guided, and strategic stages, specifically in terms of the degree to which the writer’s thinking, stance, and epistemic decisions are encoded within the prompt? Second, how is human agency, exercised through encoding decisions at the prompting stage and editorial decisions over GenAI output, associated with authorship preservation in the final text? Third, how do EFL doctoral researchers perceive and reflect on their authorship, epistemic responsibility, and critical engagement across the three conditions? The literature review situates the study within current scholarship on prompting literacy, human agency, and responsible GenAI use; the methodology section outlines the research design, instruments, and analytical design; the findings address each research question; and the discussion and conclusion draw out theoretical and pedagogical implications.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Literature Review LLMs in Academic Writing: Opportunity, Risk, and the Authorship Question The integration of LLMs into higher education writing has generated both optimism and concern. GenAI tools offer personalized learning support, reduce language barriers for multilingual writers, and expand access to academic resources (Plaatjies & Van Wyk, 2025). On the other hand, a recent synthesis of evidence from 2023 to 2025 concluded that the ability of LLMs to produce texts closely resembling human writing raises fundamental questions about the originality and authorship of student work, with the most pressing concern for educators lying in the risk that assignments will be completed in whole or in part using generative AI (Sanz-Tejeda et al., 2026). Without deliberate intentionality, writers risk blurring the boundary between their thinking and AI-generated content, surrendering rather than exercising authorship. A growing consensus in academic publishing holds that AI tools cannot qualify as authors, since they cannot take responsibility for the work submitted and, as nonlegal entities, cannot assume the accountability that authorship carries (COPE Council, 2023). Yet this consensus does not resolve the pedagogical question of how much human authorship survives AI-assisted writing. As Bozkurt (2024) argued, a balanced approach to AI integration must leverage its benefits while safeguarding against the erosion of authentic scholarly practice. The present study argues that authorship erosion depends on how writers enter AI interaction, specifically the intentionality embedded in their prompts. Prompt Engineering and Prompting Literacy in Higher Education Within this body of work, prompting has emerged as the pivotal act. Prompting mediates human intent and AI output, yet it remains under-theorized and under-assessed in higher education. A systematic review by Lee and Palmer (2025) reported that developing skills in prompt engineering can function as a critical thinking exercise, and that key concepts in effective prompting include role assignment, contextual framing, and constraint specification. The review established a prompting sophistication continuum: zero-shot prompts lack contextual detail, whereas structured prompts with defined constraints produce more aligned outputs. This continuum provides terminological and conceptual grounding for the naive, guided, and strategic prompting conditions examined here. The foundational empirical basis for this continuum was established by Zamfirescu-Pereira et al. (2023), whose study demonstrated that non-AI-expert writers approach prompting opportunistically rather than systematically, exploring prompt designs without strategic intent and producing outputs that poorly reflect their intended meaning. Their findings established that naive prompting produces misaligned outputs, while deliberate and structured approaches yield results that better reflect the writer’s intent. This distinction anchors the prompting intentionality construct developed in this study. More recent empirical work has extended this finding into educational contexts. Kim et al. (2025) analysed prompt patterns of 19 university students on AI-assisted academic writing tasks, finding that high-AI-literacy students produced descriptive, context-based prompts in collaborative interactions, while low-AI-literacy students

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

produced general, underdeveloped prompts in a more passive approach, with significant differences in writing performance across content, structure, and expression. For English language learners specifically, Tour and Zadorozhnyy (2025) demonstrated that effective AI interaction depends on the ability to contextualize requests, showing how an underdeveloped prompt produces generic output while a structured, contextually rich prompt produces output more aligned with the writer’s communicative intent. Despite this growing body of evidence, the literature offers no validated instrument for assessing prompting skill in academic writing. The prompting intentionality rubric developed for the present study and described in the Methodology was designed to address this gap. Human Agency, Epistemic Authority, and Authorship Preservation Beyond prompt quality, a second strand of research concerns agency. The relationship between AI use and writers’ sense of agency and ownership has attracted increasing empirical attention. A normative framework study argued that when prompt design requires personal discernment, contextual understanding of the AI model’s linguistic behaviour, and the capacity to produce a perceptible and deliberate effect on the output, it may qualify as a legitimate creative contribution (Ramos-Zaga, 2025). This legal argument translates directly into a pedagogical one: writers who encode their own thinking, stance, and epistemic decisions into the prompt before the AI responds create a traceable authorial record that those who prompt naively cannot. The present study operationalizes this argument empirically within an EFL academic writing context. The question of AI disclosure is inseparable from these concerns. When writers cannot account for which ideas in their final text are their own and which are AI-generated, transparent disclosure becomes impossible. Responsible AI integration requires intentional prompting that preserves authorial traceability necessary for disclosure. This study examines the relationship between prompting intentionality, authorship preservation, and ethical transparency within a single analytical framework. The Algerian EFL Context and the Research Gap These strands share a common limitation. Research on AI in academic writing remains overwhelmingly concentrated in anglophone, Western, and East Asian higher education contexts, with the Algerian EFL context remaining underrepresented in this literature. Two recent studies begin to address this gap within the Algerian context. Sebbah (2025) explored Algerian EFL students’ familiarity, use, and attitudes toward generative AI tools with 305 graduate and undergraduate students, finding that ChatGPT was the most widely used tool and that willingness to integrate GenAI was conditioned on the availability of ethical guidelines and adequate training. Boudouaia et al. (2024) conducted an experimental study with 76 Algerian undergraduate students examining ChatGPT’s impact on EFL writing, finding that AI-assisted groups outperformed control groups in writing quality. Both studies focused on

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

attitudes, perceptions, or output quality rather than on the behavioural mechanisms through which Algerian EFL writers interact with AI during the prompting act. Existing studies have tended to examine these dimensions separately rather than within a single integrated design. This fragmentation leaves limited empirical evidence addressing whether prompting intentionality functions as an authorship-preservation mechanism. The present study addresses this gap through a staged task-based process-tracing protocol applied with EFL doctoral researchers in Algeria; a population for whom questions of authorship carry relevance. Theoretical Framework This study is grounded in three complementary theoretical frameworks that together account for the cognitive, evaluative, and attitudinal dimensions of prompting intentionality in AIassisted academic writing. Writing Process Theory (Flower & Hayes, 1981) conceptualizes writing as a goal-directed cognitive activity comprising planning, translating, and reviewing. In the context of AI-assisted writing, the prompting act functions as a planning behaviour of epistemic significance. A naive prompt risks delegating aspects of planning to AI. A strategic prompt that encodes the writer’s argumentative position, assigns the LLM a bounded role, and specifies epistemic conditions complete a deliberate planning act that preserves cognitive ownership before the LLM responds. The pre-structural authorship stage corresponds to the goal-setting component identified in Flower and Hayes’s (1981) model. Critical Digital Literacy, developed within media literacy and digital education scholarship (Buckingham, 2007; Lankshear & Knobel, 2008), refers to the capacity to critically evaluate, interrogate, and take informed and responsible decisions about digital content. In the context of AI-assisted writing, it manifests as the writer’s ability to read GenAI output with a questioning eye, verify claims and sources independently, and take epistemic responsibility for every element that enters their final text. Responsible GenAI use in this study is understood as a behavioural expression of critical digital literacy. It is operationalized through three measures introduced in the Methodology: an engagement integrity composite recording how actively participants checked and revised GenAI output; a reflection section capturing participants’ own accounts of authorship; and a reference verification procedure establishing whether supplied sources were authentic. The Technology Acceptance Model (TAM) proposed by Davis (1989) holds that technology acceptance is shaped by perceived usefulness and ease of use. Participants evaluated each prompting condition using a Likert-scale battery organised into eight constructs covering agency, authorship, effort, and epistemic responsibility. TAM grounds this battery: perceived usefulness and ease of use provide the evaluative dimensions along which participants judged each condition. It also illuminates a central tension: naive prompting is easy and requires minimal cognitive investment, while strategic prompting demands planning, critical

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

engagement, and epistemic effort. The study examines whether experiencing prompting progression reshapes participants’ understanding of useful GenAI interaction. Methodology Research Design This study employs a single-institution instrumental case study (Stake, 1995) using a staged process-tracing protocol as its primary method, with reflective elicitation and embedded attitudinal measures as secondary data streams. Process tracing, adapted from cognitive writing-process research (Flower & Hayes, 1981), here captures behavioural evidence at the point of GenAI interaction, and the task design draws on task-based performance assessment (Norris, 2016). The protocol stages a single writing task into five sequential steps completed in one sitting: a human-only ideation stage (Stage 0), three successive prompting conditions using the same GenAI tool (Stages 1-3), and a final human re-authoring stage (Stage 4). The three prompting conditions constitute the study's principal comparison, while Stages 0 and 4 bracket them, establishing what the writer brought to the task before any GenAI contact and recording what survived of it afterwards. Each stage is described in full in the Staged Protocol section below. Unlike attitudinal surveys or product-based comparisons, this design treats the prompting act as the primary unit of analysis rather than relying solely on retrospective selfreport. These conditions (naive, guided, and strategic) approximate the progression observed when writers interact with AI tools, moving from minimal requests toward deliberate and structurally rich prompts (Zamfirescu-Pereira et al., 2023). By holding the task, topic, and GenAI tool constant across conditions, the design foregrounds prompting intentionality as the variable of primary interest. No claims of statistical generalizability are made. The analytical goal is transferability: generating a conceptual model sufficiently documented for researchers in comparable contexts to assess its applicability (Lincoln & Guba, 1985). Participants and Sampling Participants were doctoral candidates enrolled in the Doctoral program in the Department of English at the Teacher Training College École Normale Supérieure de Bouzaréah (ENSB), an Algerian higher education institution. Of the eighteen eligible doctoral candidates, the researcher was excluded, yielding a pool of seventeen participants. Sixteen participants completed the full protocol (94.1% response rate). The sampling approach was criterion-based purposive sampling (Patton, 2015). Participants were selected because they were active EFL academic writers whose professional practice depends on maintaining epistemic authority over their scholarly writing, making them an analytically appropriate population for this inquiry. The researcher is a member of the same doctoral cohort. This positionality is disclosed transparently and aligns with qualitative case research traditions in which insider access can enhance contextual interpretation (Merriam, 1998).

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

The Writing Task All participants completed the same writing task: an academic essay of 500 to 700 words on the topic Artificial Intelligence integration in academic writing: opportunities and risks requiring background, a rationale, and a clear argumentative stance in academic tone with reference to sources. Participants were told explicitly that the task concerned authorship, authority, and decision-making rather than writing quality or individual performance. Instrument and Procedure Data were collected through a purpose-built Google Form instrument comprising five task stages followed by post-task evaluation and reflection sections. Participants completed the task in one uninterrupted sitting of approximately 60 to 75 minutes, conducting all GenAI interaction within a single conversation thread that they were asked not to delete, reset, or restart. All sixteen participants completed all five stages. The instrument was developed by the researcher through iterative refinement and reviewed by the supervising researcher prior to deployment, serving as an expert-validation step appropriate for small-scale case-based mixedmethods designs (Creswell & Plano Clark, 2017). Given the constraints of a small purposive sample, pilot testing with cohort members was not conducted, to avoid reducing the available participant pool. Data were drawn from the protocol as follows: the three prompts (Stages 1-3) for intentionality scoring, the Stage 0 brainstorm and Stage 4 annotated outline for authorship tracing, GenAI's reference lists for verification, the conversation threads for behavioural observation, and the post-task measures for attitudinal and reflective evidence. ChatGPT was selected based on prior survey evidence indicating it as the most widely used GenAI writing tool among the target population. This supported ecological validity by minimizing interface unfamiliarity as a confounding variable. The Staged Protocol This stage operationalizes the pre-structural authorship act and serves as the baseline against which all subsequent GenAI interaction is measured. •

• •

Stage 1 required participants to prompt ChatGPT to generate the essay without adding constraints, stance, role assignment, or epistemic limits; an example prompt was supplied. Ideation necessarily preceded it since a stance formed after seeing GenAI output could no longer serve as an independent record of the writer's thinking. Stage 2 required a guided prompt including length, structure, thematic focus, and a request for sources within the same conversation thread. Stage 3 required a strategic prompt explicitly encoding the participant’s personal stance, key ideas, role assigned to the LLM, requirements for verifiable sources, and an

IAFOR Journal of Education: Technology in Education

•

Volume 14 – Issue 2– 2026

instruction to flag uncertainty rather than fabricate. Participants were instructed to paste the portion of GenAI output they found most relevant or useful at each stage. Stage 4 required participants to write a final essay outline reflecting their stance, their selected arguments, and their decisions to accept, modify, or reject GenAI-generated content. Beside each point in the outline, participants indicated in their own words whether the idea originated in their Stage 0 brainstorm, in the GenAI output, or in a combination of both. Annotation was at the level of the individual outline point and recorded inline by the participant. Participants also submitted the link to their full ChatGPT conversation thread. Prior to final submission, participants completed an external resource disclosure item indicating whether any resources beyond the designated ChatGPT thread were consulted, enabling documentation of external inputs as analysable data.

Post-Task Measures Following Stage 4, participants completed three post-task measures, all developed by the researcher for this study, as no existing instrument measured the constructs under investigation. A six-item engagement verification section measured behavioural engagement with GenAI output, recording whether participants read, revised, checked, verified, or questioned what they received. Items were answered Yes, Partially, or No, recoded as 2, 1, and 0 and summed to yield an engagement integrity composite scored out of twelve. A Likert battery rated from 1 (strongly disagree) to 5 (strongly agree) measured how participants evaluated each condition and their own role within it, across eight constructs: evaluation of naive, guided, and strategic prompting; comparative evaluation; agency and control; prompting as effort; authorship and ownership; and responsibility. The evaluation constructs operationalize the perceived usefulness and ease-of-use dimensions of TAM (Davis, 1989); the remainder operationalize human agency and epistemic authority. A reflection section of ten open-ended questions elicited participants' own accounts of agency, authorship, and epistemic responsibility, covering where they felt most in control and why, how their experience differed across conditions, whether they detected hallucinations and how, and what they took from the task overall. A copy of the full instrument is available for consultation and replication purposes at the following link: https://forms.gle/KKgSgFWKrGNvAe2J7 Data Collection and Analysis The study generated six analytically distinct data streams analysed through complementary methods within a convergent mixed methods design in which qualitative and quantitative evidence was generated in parallel and integrated at the interpretation stage (Creswell & Plano Clark, 2017).

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Prompting Intentionality Analysis The prompting intentionality rubric was developed by the researcher for this study, as no validated instrument exists for assessing prompting skill in academic writing. Its five criteria were derived from the prompting elements identified in prior work as markers of structured prompting, namely role assignment, contextual and content constraints, and constraint specification (Lee & Palmer, 2025; Zamfirescu-Pereira et al., 2023), extended with two epistemic criteria, source requirement and epistemic limit, that follow from the study's focus on authorship and epistemic authority. The rubric itself was refined iteratively and its criteria reviewed by the supervising researcher prior to coding. The criteria correspond directly to the elements participants were asked to encode at the strategic stage. Each participant’s three prompts were scored using the rubric presented in Table 1. Each criterion was scored 0 for absent or 1 for present, producing a score out of 5 per stage per participant. Coding was conducted by the researcher against the operational definitions specified in Table 1. The full corpus was then re-coded after an interval of approximately two months, and any discrepancies were reconciled by reference to the operational definitions to ensure consistent application across all prompts. Independent second coding was not feasible within the study timeline; this is acknowledged as a limitation. Table 1 Prompting Intentionality Rubric Criterion

0 = Absent

1 = Present

C1: Role Assignment

No role assigned to the GenAI

GenAI given a specific function such as language editor or content organizer

C2: Content Constraints

No constraints specified

At least two of: length, structure, tone, or thematic focus explicitly constrained

C3: Personal Stance

No authorial position included

Author’s own argumentative position on the topic explicitly stated

C4: Source Requirements

No source requirement

Request for real, verifiable, and academic sources included

C5: Epistemic Limits

No epistemic instruction

Instructions to flag uncertainty or avoid fabrication present

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Authorship Trace Analysis Each idea in the Stage 4 final outline was coded by the researcher as human-originated (present in Stage 0), AI-originated (absent from Stage 0 but present in the GenAI output), or coconstructed (derived from GenAI output but shaped using Stage 0 ideas). The sum of humanoriginated and co-constructed proportions constituted the Human Agency Preservation Ratio (HAPR) per participant, expressed as a percentage. HAPR served as the primary dependent variable and behavioural measure of authorship preservation. Reference Verification Each reference supplied by the GenAI was checked against academic databases and classified as authentic, unverifiable or fabricated, or not a citation. Coding was categorical at the level of the individual reference. Because reference counts varied widely across participants, results are reported as participant counts per stage rather than as an aggregate rate. Thread Review In addition to the form responses, the researcher reviewed participants' ChatGPT conversation threads during data collection and recorded field notes on prompting and output-handling behaviour. Of the sixteen threads, nine were available for review; the remainder had been deleted or would not load. These thread-review observations are reported as qualitative data and are not presented as independently auditable because the threads were not retained for verification. Reflexive Thematic Analysis Reflection responses were analysed using theoretically informed reflexive thematic analysis (Braun & Clarke, 2006, 2019), guided by five sensitizing constructs: human authority, prompt intentionality, agency negotiation, ethical reflexivity, and perceived progression. Participant quotations are reproduced verbatim. Quantitative Attitudinal Analysis Frequency distributions were computed for all Likert-scale items using IBM SPSS Statistics to examine response patterns across eight constructs. Descriptive statistics including means and standard deviations were computed for each construct. Construct means are reported for all eight constructs. Naive evaluation and authority, agency, and control combine positively and negatively worded items, so key items are reported alongside their means rather than the mean alone. A Friedman test was conducted, with Kendall’s W reported as the effect size to examine directional patterns across prompting conditions, appropriate for ordinal data and small sample size. A Spearman rank correlation was computed between prompting intentionality scores and the HAPR to examine the study’s central directional claim. The engagement integrity

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

composite was scored out of twelve per participant. Stage selection items were analysed using frequency distributions. Trustworthiness and Ethical Considerations Trustworthiness was addressed through triangulation across six data streams, thick description of the research context and participant profile, and reflexive positioning through transparent positionality disclosure (Lincoln & Guba, 1985). All analytical procedures and instrument items are documented; because the cohort is small and identifiable, the coded dataset is not deposited, and this constraint is acknowledged as a limitation. Claims are bounded by the case study design. Formal institutional ethical approval for the study was granted by the Head of the Doctoral Programme in accordance with ENSB institutional guidelines. Participation was voluntary and responses were anonymized. Participants confirmed informed consent prior to accessing the task. Given the peer relationship between the researcher and participants, no incentive or pressure was used to encourage participation. Data were stored securely and accessed only by researchers. Findings are reported using non-identifiable excerpts. Findings Findings are organized by research questions (RQs) and draw on six complementary data streams: prompting intentionality scores (rubric analysis), authorship trace analysis (HAPR), reference verification, engagement integrity scores, Likert-scale descriptive statistics, and reflexive thematic analysis. Points of convergence between qualitative and quantitative evidence are identified where relevant. Findings for RQ1: How Does Prompting Intentionality Develop Across the Three Stages? Prompting intentionality scores increased significantly across the three stages. The mean score rose from 0.31 (naive) to 2.56 (guided) to 3.56 (strategic), indicating a marked increase from entry to the strategic condition. A Friedman test confirmed a significant effect of stage, χ² (2) = 25.20, p < .001, with a large effect size (Kendall's W = 0.79). Bonferroni-corrected Wilcoxon signed-rank comparisons showed significant increases at every transition, including guided to strategic (p = .045). All sixteen participants progressed from naive to strategic. At the naive stage, encoding was minimal. Fifteen of 16 participants (93.8%) scored 0 or 1. Prompts averaged twelve words and typically contained no personal stance, role assignment, source requirement, or epistemic limit, functioning as task submissions rather than authorial directives (for example, “Write an essay about the use of AI tools in academic writing”). Source and epistemic language were entirely absent at this stage. No participant included instructions to verify sources, flag uncertainty, or avoid fabrication (C4 = 0 and C5 = 0 for all 16). Reference verification confirmed that no references were produced.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

The guided stage produced the largest single-stage increase. Content constraints became nearuniversal, with 15 of 16 participants (93.8%) specifying word count, structure, thematic focus, or tone. Source and epistemic requirements also increased substantially when the guided task demanded them: eight (50.0%) participants specified source requirements, and 11 (68.8%) included epistemic instructions (C4 = 8, C5 = 11). Mean prompt length increased to approximately sixty-nine words. At the strategic stage, the defining development was explicit personal stance, encoded by 15 of 16 participants (93.8%) through first-person argumentative positioning, for example, “I argue that...”. Mean prompt length rose to approximately 110 words, and the maximum rubric score of 5 out of five was reached by one participant at the guided stage and by three participants at the strategic stage. The criteria differed in their uptake: personal stance, content constraints, and epistemic limits were each encoded by most participants, whereas role assignment remained the least frequent even at the strategic stage, at 8 of 16 (50.0%). The lower content-constraint count at the strategic stage relative to the guided stage (11 versus 15) reflects how participants specified constraints rather than a decline in intentionality. Five participants who had specified constraints at the guided stage did not restate them at the strategic stage, while one participant added them. Three of the five (P03, P10, P16) built on prior turns with formulations such as "do the same but"; the remaining two encoded stance and content without form-level specifications. Because coding was conducted at the level of the individual prompt, constraints carried over implicitly were scored as absent. (see Table 2) Table 2 Prompting Intentionality Scores Across Stages (n = 16) PID

Naive

Guided

Strategic

Total

P01

P02

P03

P04

P05

P06

P07

P08

P09

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

PID

Naive

Guided

Strategic

Total

P10

P11

P12

P13

P14

P15

P16

Mean

0.31

2.56

3.56

0.60

1.26

0.96

Note. Each prompt was scored against five criteria (Table 1), one point per criterion; stage scores therefore range from 0 to 5 and total scores from 0 to 15.

Table 3 Criterion-Level Presence Across Prompting Stages (n = 16) Criterion

Naive n (%)

Guided n (%)

Strategic n (%)

C1: Role Assignment

0 (0.0%)

2 (12.5%)

8 (50.0%)

C2: Content Constraints

4 (25.0%)

15 (93.8%)

11 (68.8%)

C3: Personal Stance

1 (6.3%)

5 (31.3%)

15 (93.8%)

C4: Source Requirements

0 (0.0%)

8 (50.0%)

11 (68.8%)

C5: Epistemic Limits

0 (0.0%)

11 (68.8%)

12 (75.0%)

Table 3 reveals a clear ordering in how readily writers adopted each dimension of intentional prompting. Content constraint specification was the most naturalistic entry point, used by 93.8% of participants at the guided stage. Personal stance encoding was the defining development of the strategic stage, adopted by 93.8%. Source requirements and epistemic limits emerged principally once the guided and strategic tasks called for them, reaching 68.8% and 75.0% respectively at the strategic stage, while role assignment remained the least adopted criterion throughout, reaching 50.0% at the strategic stage.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Figure 1 Criterion-Level Prompting Intentionality Trajectories Across the Three Stages (n = 16)

Two participants illustrate contrasting high-development trajectories. P07 achieved the highest total prompting intentionality score in the dataset (10 of 15), reaching the maximum of five out of five at both the guided and strategic stages; the prompts specified a bounded role, structural constraints, a stated stance, source requirements, and explicit epistemic instructions. P09 demonstrated a different pattern, encoding at the strategic stage a precise personal position that simultaneously satisfied several criteria. P09 summarized the progression across the three stages as follows: "Naive: ChatGPT decided. I reacted. Guided: I added requirements. Better but not quite mine. Strategic: I decided everything. I controlled." (see Figure 1) Findings for RQ2: How Is Human Agency Associated with Authorship Preservation? The authorship trace analysis yielded a mean HAPR of 79.4% (n = 16), indicating that participants retained, on average, approximately four-fifths of their initial thinking in the final outline. Human-originated content accounted for 52.5% of final outlines, co-constructed content for 26.9%, and AI-originated content for 20.6%. Table 4 Human Agency Preservation Ratio by Participant PID

AI%

CO%

HAPR%

Category

P07

High

P08

Moderate

P09

High

P10

Moderate

P11

High

P12

Moderate

P13

High

P14

High

P15

Moderate

P16

High

Mean

52.5

20.6

26.9

79.4

20.49

12.76

14.01

12.76

Note. High HAPR ≥80%: 10 participants (62.5%). Moderate 65-79%: 4 (25.0%). Low <65%: 2 (12.5%).

These thresholds are descriptive rather than diagnostic: the ≥80% band identifies outlines in which human and co-constructed content clearly predominated, the 65–79% band those with substantial but more mixed human contribution, and the <65% band those where AI-originated content approached or exceeded one third. The cut-offs are used only to organize discussion of the cases and carry no external standardized meaning. Prompting intentionality was positively associated with authorship preservation. Strategicstage prompting intentionality was associated with HAPR (rs = .62, p = .011, 95% CI [.12, .91]), and total prompting intentionality showed a stronger association (rs = .76, p < .001, 95% CI [.42, .92]). Both associations are significant, though the wide confidence intervals reflect the small sample (N = 16) and warrant cautious interpretation. (see Table 4)

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Figure 2 Prompting Intentionality and Human Agency Preservation Ratio by Participant, Ordered by Intentionality (n = 16)

Three cases illustrate the relationship. P07 demonstrated the clearest alignment between high prompting intentionality and high authorship preservation, producing the highest total intentionality score in the dataset alongside a HAPR of 95%, with a Stage 4 outline that explicitly annotated the origin of each idea. P13 showed comparable alignment, achieving a HAPR of 95% with a final outline that remained nearly identical to the Stage 0 brainstorm, indicating sustained authorial control. In contrast, P02 recorded the lowest HAPR (50%). The final outline incorporated substantially more AI-generated content than that of any other participant, while prompting intentionality increased only modestly across stages (naive = 0, guided = 2, strategic = 3). P02 reflected feeling largely absent as a writer and reported delegating responsibility to the GenAI tool. (see Figure 2) Reference verification provided an additional behavioural dimension. At the naive stage, ChatGPT produced zero references across all sixteen participants. References appeared at the guided and strategic stages almost exclusively for participants who had requested sources in their prompts, consistent with source specification functioning as a driver of source-generating behaviour. Two probable hallucinations were identified: P13 received a reference to an unverifiable paper by Algerian authors at the guided stage, P12 received keyword labels rather than actual citations at both stages. Thread review, conducted for the nine available conversation threads, indicated distinct outputhandling behaviours. Several participants (P02, P03, P05, P13) carried the GenAI output across unchanged. Three participants engaged in reference-management follow-ups beyond the required stages: P09 asked GenAI directly, "Are you sure about all sources?", after the guided and strategic stages and requested a separate reference list; P16 asked the GenAI to organize references into a separate list after the guided stage; and P11 asked, as a final prompt, for GenAI to combine its suggestions with the Stage 0 material.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

The engagement integrity composite yielded a mean score of 6.56 out of twelve, indicating moderate engagement with GenAI output. The item measuring active revision of GenAI output received the lowest responses, with ten of the sixteen participants answering No, indicating limited exercise of agency at the output-selection stage. Reflections confirmed this unevenness: while several participants reported verifying sources, others stated plainly that they did not check the output, one noting simply, “honestly, I didn’t check” (P08). Findings for RQ3: How Do Participants Perceive and Reflect on Authorship, Agency, and Epistemic Responsibility? Likert-scale results and thematic analysis converged around five patterns describing participants’ psychological and epistemic experiences across the staged protocol. Table 5 Likert Construct Means, Ranked (n = 16, scale 1-5) Rank

Construct

Mean

Key Items

PE: Prompting as Effort

4.42

PE1=4.50, PE2=4.50

RE: Responsibility and Ethics

4.34

RE2=4.44

CE: Comparative Evaluation

4.21

CE3=4.31

SE: Strategic Evaluation

4.06

SE1=4.44, SE4=4.31

GE: Guided Evaluation

3.99

GE4=4.12

AAC: Authority Agency Control

3.86

AAC8=4.44

AO: Authorship Ownership

3.73

AO1=3.25

NE: Naive Evaluation

3.66

NE2=4.19

Pattern 1: Prompting as Deliberate Cognitive Labor. The Prompting as Effort construct yielded the highest mean (4.42). The two highest-scoring items (both 4.50) indicated that output quality depended on prompt formulation and that prompting required conscious decision-making. All participants agreed or strongly agreed on the first item, and fifteen of sixteen on the second. Within this dataset, participants experienced prompting as cognitively demanding authorial work, aligning with the behavioural findings under RQ1. Pattern 2: Strategic Prompting as Highest Perceived Agency concerns the progressive recovery of authorial identity. All sixteen participants identified the strategic stage as the condition of strongest authorship. CE3, the item measuring progressive refinement of agency across the

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

staged progression, received agreement or strong agreement from all sixteen participants (mean 4.31). Stage-selection items for highest control, most effort, and preferred mode showed consistent directional preference toward strategic prompting. Several participants used metaphors such as editor or director, suggesting a shift in perceived role from content generator to supervisory authority over AI-mediated production. Pattern 3: Naive Prompting as Reduced Agency. Nearly all participants identified the naive stage as the condition of strongest GenAI influence and weakest authorship. P02 wrote that they were absent from the production as writers. P05 said they gave the GenAI tool green lights to do whatever. P13 described the naive stage as making them feel unethical, stating that the content was 100% not mine. P02 described their own shift over the task from a "careless, irresponsible simple-task performer to a conscious, responsible" writer. Within this sample, naive prompting was retrospectively framed as diminished agency and weakened epistemic responsibility. Pattern 4: Emergent Verification Urge and Epistemic Awareness. This pattern concerns an internal urge to verify emerging alongside intentional prompting. P02 described feeling an urge to verify as their prompts became more specific, a behaviour the instrument did not require. Thread review indicated a comparable pattern for P09, who questioned ChatGPT about the reliability of its sources during the task. While participants exercised agency during prompt encoding, self-reports indicated weaker agency at output selection: ten of sixteen reported not revising the AI output. This asymmetry between encoding and selection was not explicitly scaffolded in the instrument design. Pattern 5: Ownership Remains Qualified. The Authorship Ownership construct yielded the second-lowest mean (3.73). AO1 asking (whether the final text was fully their own) scored 3.25, the lowest individual item in the instrument. P03 and P04 both explicitly qualified their sense of authorship, with P03 stating they did not feel 100% the author and P04 noting that the essay remained AI-dependent even when prompts were controlled. The staged structure of the instrument itself was perceived as pedagogically significant. AAC8, measuring whether the staged structure increased authorial awareness, scored 4.44. P08 articulated this most fully: "LLMs are exceptional at structural organization and stylistic refinement, but the human element is still required to define the purpose, stance, and intellectual direction of the work." P01 drew the most ethically grounded conclusion: “I must verify AI claims as part of my authorial responsibility as an author.” (see Table 5) Discussion This study investigated how prompting intentionality is associated with the preservation of human authorship in GenAI-assisted academic writing. The findings across the three research questions suggest a consistent relationship operating across behavioural, perceptual, and attitudinal dimensions.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Prompting as a Planning Act: Extending Writing Process Theory The finding that prompting intentionality scores increased significantly across the naive, guided, and strategic stages, rising from a mean of 0.31 to 3.56, is consistent with the cognitive process model of writing proposed by Flower and Hayes (1981), which positions planning as the foundational cognitive act that precedes and shapes all subsequent writing behaviour. This study extends that model into the AI-mediated writing context by suggesting that the prompting act functions as a planning behaviour with epistemic consequences. A naive prompt may signal that planning has been delegated to the LLM; a strategic prompt encodes the writer’s reasoning into the interaction before the GenAI produces a single word. Stage 0 operationalizes the moment Flower and Hayes’s model designates as foundational, goal-setting and idea generation before any text is produced; the prompting act that follows may preserve or weaken that foundation. The pattern across criteria adds an important qualification. Rather than a uniform rise, the data shows a clear ordering in which some dimensions of authorial encoding were readily adopted and others resisted. Personal stance, content constraints, and epistemic limits were each encoded by most participants at the strategic stage, whereas role assignment remained the least frequent, reached by only half. Role assignment requires the writer to specify not what to say but how the tool should function and appears to demand a distinct and less intuitive form of planning. This suggests that planning in AI-mediated writing involves a metacognitive layer beyond conventional writing planning: writers must shape not only their own ideas but the conditions of the GenAI interaction itself. Authorship Preservation, Prior Knowledge, and Writer Typology The mean HAPR of 79.4% is both encouraging and analytically complex. On the surface it suggests that participants preserved most of their intellectual DNA in the final text. However, the mechanisms underlying this preservation differed across participants. For participants like P13 and P07, high HAPR was achieved through a combination of strong pre-AI ideation at Stage 0 and high prompting intentionality across the guided and strategic stages. These participants entered the GenAI interaction already knowing what they wanted to argue and used increasingly intentional prompts to channel the tool within their own intellectual framework. Their experience aligns with Zamfirescu-Pereira et al. (2023), who found that prompting quality shapes the degree to which output reflects writers’ intent. For participants like P01 and P10, HAPR rested on the richness of the Stage 0 brainstorm itself. This authorship trace pattern, together with thread-review observations, may also point to an emergent writer typology, though the small sample means this remains a hypothesis for future testing. Thread review indicated distinct approaches to the early stages: some participants appeared to use the naive stage as thematic orientation before consolidating their own stance, others treated it as verification against pre-existing ideas, and a third group incorporated naive output more directly into later stages. For the first two groups, the early stages functioned less

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

as content sources than as cognitive prompts for clarifying their own positions, suggesting that the naive and guided stages may retain authorial value when writers bring sufficient prior knowledge to evaluate the LLM output. This finding cautions against dismissing naive prompting as simply harmful. This observation extends the findings of Kim et al. (2025), who demonstrated that AI literacy shapes prompting quality, by suggesting that domain knowledge and writers' cognitive stance may also influence authorship outcomes in AI-assisted writing. Writers with strong prior knowledge may preserve substantial authorship despite naive prompting, whereas those with more limited prior knowledge may be more susceptible to GenAI influence because the tool fills a larger cognitive gap. P02 illustrates the contrasting case. Although the participant generated genuine ideas at Stage 0, these were only partially encoded into subsequent prompts. The participant also described an absence from the writing process; a pattern reflected in the limited progression of prompting across stages and the lowest HAPR in the dataset. This case suggests that limited prompting intentionality may be associated with greater reliance on AI-generated content. The engagement integrity data adds a further behavioural dimension to this pattern. The finding that ten of the sixteen participants reported not actively revising GenAI output suggests that passive acceptance of GenAI output is a default behaviour that strategic prompting disrupts but does not eliminate across all participants. Encoding one’s own thinking into a prompt appeared insufficient on its own to ensure active editorial engagement with subsequent GenAI output. A further pattern in the authorship-trace data merits future attention: mean human-only content, prior to any GenAI contribution, averaged approximately 52.5%, well below the composite HAPR, and in several cases fell at or below one third. The present sample is too small to interpret this distribution, which a larger study should examine directly. This qualified picture matters. Participants preserved substantial authorship, yet several withheld full ownerships even at the strategic stage, with P03 and P04 stating they did not feel fully the authors of the final text. The pattern resists both the technophobic conclusion that GenAI inevitably destroys authorship and the technophilic claim that strategic prompting resolves the authorship question entirely. Prompting intentionality increased authorial control without fully restoring ownership to the pre-AI baseline. Critical Digital Literacy in Action Some of the most analytically significant qualitative findings concern how critical digital literacy emerged behaviourally through intentional prompting practices. P09 provides the clearest example. Thread review indicated that this participant questioned the LLM directly about the reliability of its sources during the task and requested a separate reference list, and in their reflection reported verifying sources against previously known references rather than accepting them uncritically. These behaviours express the epistemic

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

vigilance that Buckingham (2007) and Lankshear and Knobel (2008) identify as the core capacity of critical digital literacy. They were not required by the instrument and appeared to emerge through more deliberate engagement with the GenAI interaction. P02 articulated the mechanism behind this emergence precisely: as the instructions and criteria in their prompts became more specific, they felt an internal urge to verify the output. Verification appeared to become less procedural and more connected to participants’ sense of authorial responsibility. This might be interpreted as the more writers invest in the prompt, the more they attend to whether the GenAI output reflects that investment faithfully. The ethical vocabulary voiced by several participants strengthens this interpretation. P13 described naive prompting as unethical; P02 described their own shift from a "careless, irresponsible" performer to a "conscious, responsible" writer; and P01 articulated this role shift directly, writing that across the stages "my job moved from writing to paraphrasing and verifying sources," and that verifying AI claims formed part of their authorial responsibility. These concerns arose without explicit prompting from the instrument. By making the contrast between naive and strategic prompting experientially vivid, the staged protocol appeared to function as a reflective pedagogical experience. Strategic prompting also creates a traceable record of the writer’s intellectual DNA, making transparent AI disclosure both feasible and natural. Technology Acceptance and the Redefinition of Perceived Usefulness The Prompting as Effort construct yielded the highest mean score in the Likert battery at 4.42, and the two highest-scoring items in the entire instrument both concerned the relationship between prompting effort and output quality. This self-report finding is analytically important in relation to the TAM (Davis, 1989), which proposes that perceived ease of use and perceived usefulness are the two primary drivers of technology acceptance. Naive prompting scores high on perceived ease of use, requiring minimal cognitive investment while producing immediate output. If ease of use were the dominant criterion for technology acceptance in writing contexts, naive prompting would be the rational default and strategic prompting would be rejected as unnecessarily burdensome. The findings run counter to this expectation. The fact that 93.8% of participants would choose strategic prompting if selecting one mode after experiencing all three conditions suggests that the staged protocol shifted their understanding of what useful GenAI interaction means. Strategic prompting is more effortful, but participants experienced that effort as meaningful rather than burdensome, because it produced output more aligned with their own thinking and generated a stronger sense of authorial ownership. These findings suggest an expanded interpretation of perceived usefulness. In GenAI-assisted academic writing, usefulness may depend not only on how quickly a tool produces text but also on how faithfully the output reflects the writer's own reasoning. Accordingly, TAM-based frameworks for evaluating GenAI adoption may benefit from incorporating authorial agency

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

as an additional dimension of perceived usefulness alongside conventional productivity-based measures. The Agentic Prompting Loop: An Emergent Conceptual Pattern Across the behavioural, perceptual, and attitudinal findings, a consistent three-dimensional pattern emerges. As prompting intentionality increases across the three stages, epistemic authority shifts progressively back toward the human writer, and authorship preservation correspondingly increases. These three dimensions, prompting intentionality, epistemic authority, and authorship preservation, appeared to co-vary across the staged progression. The findings suggest that human agency in GenAI-assisted writing operates across two sequential moments. The first occurs at the prompting stage, where writers decide what ideas, stance, and epistemic requirements to encode before the GenAI tool responds. The second occurs at the output stage, where writers evaluate, select, modify, or reject AI-generated content. Within this study, the first form of agency was more consistently developed because participants received explicit prompting scaffolds across the staged progression. By contrast, output-level agency appeared less consistently exercised, as reflected in the self-reported revision data reported above. This suggests that intentional prompting alone may not ensure critical editorial engagement with GenAI output, and that output-level agency may require explicit instructional support. The HAPR, therefore, reflects not only what writers encoded into prompts but also how they negotiated AI-generated content during synthesis. (see Figure 3) This co-movement is tentatively named the Agentic Prompting Loop: a conceptual pattern in which prompting intentionality, epistemic authority, and authorship preservation operate as interdependent and mutually reinforcing dimensions of human agency in AI-assisted writing. It is proposed as an exploratory conceptual pattern requiring examination and refinement across larger and more diverse contexts.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Figure 3 The Agentic Prompting Loop: an emergent conceptual pattern (n = 16, exploratory)

Pedagogical Implications and Recommendations Three recommendations emerge from this study, addressed to educators, researchers, and institutional policymakers working at the intersection of GenAI integration and writing pedagogy. First, prompting literacy instruction in higher education should explicitly address epistemic limit specification. Criterion C5, instructing the GenAI tool to flag uncertainty or avoid fabrication, was entirely absent at the naive stage and emerged principally once the guided and strategic tasks explicitly called for it. This pattern suggests that epistemically responsible prompting may not emerge through general GenAI use alone but depends on explicit instructional scaffolding. Second, the staged protocol developed in this study may offer a replicable pedagogical approach for academic writing instruction across disciplines and language contexts. Participants consistently identified strategic prompting as the condition associated with the strongest sense of authorship, and most preferred it for future academic writing. This pattern suggests that experiential comparison across prompting conditions may support both

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

behavioural and attitudinal development in GenAI-assisted writing practices. Future implementations should consider administering the writing task and evaluative components in separate sessions to reduce task fatigue. Third, institutional AI disclosure frameworks may benefit from complementing detectionbased approaches with authorship trace models. The Stage 4 annotation procedure developed in this study, in which writers classify ideas as human-originated, AI-originated, or coconstructed, offers a transparent and pedagogically transferable approach to documenting AIassisted writing processes. Embedding idea-origin documentation within the writing process may strengthen reflective authorship practices while supporting transparent disclosure. Institutions in Algeria and comparable higher education contexts may therefore consider integrating structured authorship-tracing procedures into assessment and submission guidelines. Limitations Several limitations must be acknowledged. The sample of 16 doctoral researchers from a single department and institution in Algeria reflects a theoretically bounded case-study design rather than an attempt at broad representation: this cohort represents the only group of active doctoral researchers at the Teacher Training College ENSB, making it a theoretically relevant population for this inquiry. Nevertheless, the small sample size limits transferability, and future research should examine whether the patterns observed here replicate across disciplines, institutions, language contexts and other countries. The prompting intentionality rubric was developed and scored by the researcher, with the full corpus re-coded after a two-month interval and discrepancies reconciled against the operational definitions. Independent second coding was not feasible within the study timeline; future studies should establish formal inter-rater reliability procedures across the full dataset. Because thread review depended on participant-retained conversations, it was not available for every participant, and the threads were not preserved for independent audit. Future instrument designs should require thread submission and archiving prior to form closure. The authorship trace analysis involved interpretive judgments about idea origin that cannot be fully standardized. Participants who entered Stage 0 with substantial prior reading or topic familiarity, as observed in one case, may have produced stronger initial ideation before interacting with GenAI. Future studies should include a Stage 0 verification mechanism to distinguish independent brainstorming from research-informed brainstorming. The study operationalized human agency more explicitly at the prompting stage than at the output evaluation stage. Although participants were instructed to paste the portion of GenAI output they considered relevant, they were not explicitly scaffolded in critical selection, revision, or rejection strategies, and self-reports indicated that many did not revise the output.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Future designs should incorporate explicit instructional and analytical procedures targeting post-output editorial agency alongside prompting intentionality. A design consideration concerns the overlap between the task instructions and the scoring criteria. Because the naive task directed participants not to encode constraints and the guided and strategic tasks directed them toward some rubric criteria, scores at all three stages reflect scaffolded rather than spontaneous encoding. The findings are therefore reported as responses to staged instruction rather than as evidence of unprompted behaviour. At the criterion level, the observed ordering is interpreted as distinguishing dimensions that writers adopted readily under instruction from those they resisted even when prompted, most notably role assignment. Finally, the integrated single-session format, which combined the writing task and attitudinal battery within one sitting of 60 to 75 minutes, may have introduced task fatigue in later stages. The instrument was designed this way by necessity in the absence of a dedicated classroom setting. Future implementations should administer the writing task and evaluative battery in separate sessions, consistent with the instrument’s original two-component design. Conclusion This study investigated a question at the intersection of writing pedagogy, digital ethics, and human cognition: when doctoral researchers use GenAI tools in academic writing, how is prompting intentionality associated with authorship preservation? The findings across sixteen EFL doctoral researchers at ENSB suggest that prompting intentionality plays a meaningful role in authorship preservation, and that the mechanism through which authorship is preserved or surrendered is the degree of intentionality the writer brings to the prompting act itself. Three findings stand as the primary contributions of the study. First, source and epistemic language was absent at the naive stage across all sixteen participants, suggesting that naive prompting operated as an epistemically unguarded form of interaction and that prompting practices may shape whether hallucinated content is critically interrogated during the writing process. Second, all sixteen participants identified strategic prompting as the stage of strongest authorship, suggesting that the experience of intentional prompting is not only behaviourally productive but associated with a stronger sense of authorial presence. Third, prompting intentionality was positively associated with the HAPR (total intentionality rs = .76, p < .001), providing preliminary behavioural evidence of a relationship between intentional prompt encoding and authorship preservation in the final text. Across these findings, an emergent conceptual pattern presents itself: the Agentic Prompting Loop, in which prompting intentionality, epistemic authority, and authorship preservation operate as interdependent and mutually reinforcing dimensions of human agency in AI-assisted writing. This pattern is proposed as an empirically grounded beginning that future research in larger and more diverse contexts should examine, refine, and test. The question animating this study is not whether writers should use GenAI tools. Rather, as their use becomes increasingly widespread in academic contexts, the more pressing question is

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

whether writers can remain the authors of their own texts while doing so in ways that are transparent, accountable, and epistemically defensible. The findings suggest that they can, but not without deliberate effort, explicit instruction in prompting literacy, and institutional scaffolding that treats the pre-structural authorship stage as the foundation on which responsible AI-assisted writing can be built. More broadly, the study suggests that writing instruction in the age of GenAI should place greater emphasis on prompting literacy, epistemic responsibility, and transparent authorship practices.

Acknowledgements The authors express sincere gratitude to the doctoral researchers at ENSB, whose participation, intellectual engagement, and candid reflections made this study possible. Their willingness to examine their own authorial practices with honesty and rigor enriched the research process. The author also acknowledges the support of her doctoral supervisor throughout the research and writing process. Declaration of Generative AI and AI-assisted Technologies The authors would like to acknowledge the use of artificial intelligence (AI) tools in the preparation of this manuscript. The following tools: ChatGPT, Claude, Grammarly, and QuillBot, were used for these specific purposes only: language editing, proofreading, and stylistic refinement. AI-edited text was thoroughly reviewed and revised by the authors to ensure accuracy, clarity, and adherence to academic standards. AI tools were not used for other purposes, including data generation, data analysis, methodology, or interpretation of findings or conclusions. The authors take full responsibility for all aspects of the final manuscript.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

References Boudouaia, A., Mouas, S., & Kouider, B. (2024). A study on ChatGPT-4 as an innovative approach to enhancing English as a foreign language writing learning. Journal of Educational Computing Research, 62(6), 1509–1537. https://doi.org/10.1177/07356331241247465 Bozkurt, A. (2024). Tell me your prompts and I will make them true: The alchemy of prompt engineering and generative AI. Open Praxis, 16(2), 111–118. https://doi.org/10.55982/openpraxis.16.2.661 Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101. https://doi.org/10.1191/1478088706qp063oa Braun, V., & Clarke, V. (2019). Reflecting on reflexive thematic analysis. Qualitative Research in Sport, Exercise and Health, 11(4), 589–597. https://doi.org/10.1080/2159676X.2019.1628806 Buckingham, D. (2007). Digital media literacies: Rethinking media education in the age of the Internet. Research in Comparative and International Education, 2(1), 43–55. https://doi.org/10.2304/rcie.2007.2.1.4 3 COPE Council. (2023). COPE position - Authorship and AI-English. https://doi.org/10.24318/cCVRZBms Creswell, J. W., & Plano Clark, V. L. (2017). Designing and conducting mixed methods research (3rd ed.). SAGE. Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319–340. https://doi.org/10.2307/249008 Flower, L., & Hayes, J. R. (1981). A cognitive process theory of writing. College Composition and Communication, 32(4), 365–387. https://doi.org/10.2307/356600 Kim, J., Yu, S., Lee, S. S., & Detrick, R. (2025). Students’ prompt patterns and its effects in AI-assisted academic writing: Focusing on students’ level of AI literacy. Journal of Research on Technology in Education, 58, 638–655. https://doi.org/10.1080/15391523.2025.2456043 Lankshear, C., & Knobel, M. (2008). Digital literacies: Concepts, policies and practices. Peter Lang. Lee, D., & Palmer, E. (2025). Prompt engineering in higher education: A systematic review to help inform curricula. International Journal of Educational Technology in Higher Education, 22, 7. https://doi.org/10.1186/s41239-025-00503-7 Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. SAGE. Merriam, S. B. (1998). Qualitative research and case study applications in education. Jossey-Bass. Norris, J. M. (2016). Current uses for task-based language assessment. Annual Review of Applied Linguistics, 36, 230–244. https://doi.org/10.1017/S0267190516000027 Patton, M. Q. (2015). Qualitative research and evaluation methods (4th ed.). SAGE.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Plaatjies, B., & Van Wyk, M. (2025). Prompt literacy as an enhancer of students’ academic writing in higher education institutions: A systematic literature review. Journal of Teaching and Learning, 19(4), 114–134. https://doi.org/10.26522/jtl.v19i4.4387 Ramos-Zaga, F. A. (2025). Reconceptualizing human authorship in the age of generative AI: A normative framework for copyright thresholds. Laws, 14(6), 84. https://doi.org/10.3390/laws14060084 Sanz-Tejeda, A., Domínguez-Oller, J. C., Baldaquí-Escandell, J. M., Gómez-Díaz, R., & García-Rodríguez, A. (2026). The impact of generative AI on academic reading and writing: A synthesis of recent evidence (2023–2025). Frontiers in Education, 10, 1711718. https://doi.org/10.3389/feduc.2025.1711718 Sebbah, L. (2025). Exploring Algerian EFL students’ familiarity, use and attitudes towards generative artificial intelligence tools in education. Journal of Languages and Translation, 5(1), 1–21. https://doi.org/10.70204/jlt.v5i1.219 Stake, R. E. (1995). The art of case study research. SAGE. Tour, E., & Zadorozhnyy, A. (2025). Conceptualizing and operationalizing prompt literacy for English language learners. Journal of Adolescent & Adult Literacy, 69(3), e70020. https://doi.org/10.1002/jaal.70020 Zamfirescu-Pereira, J. D., Wong, R. Y., Hartmann, B., & Yang, Q. (2023). Why Johnny can’t prompt: How non-AI experts try (and fail) to design LLM prompts. Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems, 1–21. https://doi.org/10.1145/3544548.3581388

Corresponding author: Rym Ladjal Institutional Email: rym.ladjal@ensb.dz

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Applying Gamification to Undergraduate Accounting Education: A MixedMethods Study Tialei Scanlan Brigham Young University–Hawaii, United States Spencer Scanlan Brigham Young University–Hawaii, United States

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Abstract Gamification is increasingly being adopted across industries including education, fitness, social media, crowdsourcing, and corporate environments due to its potential to enhance motivation, engagement and performance. This study included a three group, quasiexperimental design with focus groups. The purpose of this mixed-methods study was to understand the effect of badging (a gamification element) in an undergraduate accounting course through the lens of Self-Determination Theory. The quantitative results found that motivation remained high during a period often associated with declining motivation or the mid-semester slump, but no significant differences were found in motivation between the beginning and the end of the intervention. Although students reported positive outcomes such as chunking material, tracking progress, a sense of learning autonomy, and feedback/recognition, behavioral engagement and academic performance showed no significant differences across the groups. At the end of the intervention, males in the badge experimental groups exhibited higher intrinsic motivation than males in the leaderboard only (no badge) group. Conversely, females in the leaderboard only (no badge) group exhibited higher intrinsic motivation than in the badge experimental groups. Further, females showed significantly higher behavioral engagement in terms of pageviews. The gamification intervention evoked both positive and negative outcomes in terms of motivation with some students reporting high intrinsic motivation and integrated regulation while others reporting high extrinsic regulation or amotivation. These contrasting experiences may have offset one another when averaged across the group, contributing to non-significant quantitative results for the gamification intervention. Implications for practice are discussed for students, accounting teachers, and decision makers. Keywords: academic performance, leaderboards, motivation

badges,

behavioral

engagement,

gamification,

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Student engagement is one of the most significant challenges facing instructors in higher education today, as it plays a critical role in persistence and academic success (Veiga et al., 2026). One suggestion is the use of technology to present novel types of learning activities for students to engage in deeper learning. A method of engagement that has gained attention is gamification, or the use of game elements in a non-game context (Alvi, 2025; Cainas & Kralik, 2024; Chugh & Turnbull, 2023; Oliveira et al., 2023). Gamification and, more specifically, badges and leaderboards, have been trending topics to boost engagement and enhance user activity (Goulding et al., 2024; Tahir et al., 2022; Velazquez-Garcia et al., 2024). Badges are a representation of achievement or a mark of completing a task, often a complex one (Almeida et al., 2023). Understanding if gamification and more specifically badging really works and who it works for, is a practical and relevant issue for educators to make wise decisions when delivering content. This study focuses on the gamification elements of badges and leaderboards within an undergraduate accounting course in a traditional, face-to-face environment context. The researcher paired badges and leaderboards in the experimental group because they are among the most commonly used gamification elements, alongside points, levels, and progress bars (Dichev & Dicheva, 2017; Tahir et al., 2022). Arguably, these two gamification elements have had the most success in terms of motivation and behavioral gains and students often prefer the inclusion of multiple gamification features (Haruna et al., 2018; Ortega-Arranz et al., 2019; Rodrigues et al., 2022). Some previous studies have awarded both participation-based and skill-based badges simultaneously within a course setting, making it difficult to determine differences between these badge types (Kyewski & Krämer, 2018; Ortega-Arranz et al., 2019; Tahir et al., 2022). In the present study, leaderboards were integrated across all three groups. The control group included leaderboards but no badges, allowing the researcher to isolate and examine differences between participation-based and skill-based badges. To strengthen the study design, the researcher controlled for game preferences and cumulative GPA while examining motivation, behavioral engagement, and academic performance across gender within an accounting course. Accounting students may respond differently to gamification than students in other disciplines because accounting is typically characterized by structured content and procedural problem-solving, whereas other disciplines may place greater emphasis on creativity and exploration. These differences may influence how students engage with gamified elements. Literature Review and Research Questions Although research on gamification is generally positive (Abadi et al., 2022; Kim & Castelli, 2021; Velazquez-Garcia et al., 2024), research on the use of badging within higher education contexts has been heterogeneously applied (e.g., Abadi et al., 2022; Chugh & Turnbull, 2023; Oliveira et al., 2023; Tahir et al., 2022). Across higher education settings, badges have been used for a variety of purposes, including supporting both optional and compulsory assignments (Ortega-Arranz et al., 2019), recognizing completion of compulsory course components such as modules (Velazquez-Garcia et al., 2024), promoting sustained engagement with learning

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

tasks (Tahir et al., 2022), and serving as digital credentials that signify demonstrated competence in specific skills (Steenkamp et al., 2024). Gamification has been applied in many small-scale educational contexts with insufficient statistical evidence, especially regarding learning performance (e.g., Sasmaz et al., 2025). As such, best practices for the most effective implementation of badges are still being compiled (Velazquez-Garcia et al., 2024). This study focused on one aspect of behavioral engagement as defined by Fredricks et al., (2004) which includes: participation and involvement in learning and academic tasks. Students who are motivated and complete self-determined actions produce higher-quality behavior (Deci & Ryan, 1994; Kocsis & Molnar, 2025; Laranjeira & Teixeira, 2025). The framework of this study was based on the self-determination theory (Deci & Ryan, 1985). Selfdetermination theory is a broad framework of motivation that proposes three needs–autonomy, competence, and relatedness–which have an impact on well-being within a setting. Competence is the perceived extent of one's own actions as the cause of desired consequences in one's environment; autonomy is the need for people to experience their behaviors as selfdetermined; and relatedness is the need for people to feel close to others (Deci & Ryan, 1985). These three needs foster motivation and motivation can be measured in a scale that includes: Intrinsic Motivation, Integrated Regulation, Extrinsic Regulation and Amotivation (Guay et al., 2001). Intrinsic motivation is the product of self-determined behaviors, as it is autotelic in nature, self-initiated, or done out of volition (Deci & Ryan, 1994). Integrated regulation is the most determined, autonomous, and mature regulatory style of extrinsic motivation (Ryan & Deci, 2000). Extrinsic motivation can become internalized, where individuals integrate the activity within themselves, and there is congruence between the individual’s goals and the task, even though the regulation is generated externally (Deci et al., 1994). Amotivation refers to a lack of motivation. Research Questions The following quantitative and qualitative questions guided this study. Q1: Is there a difference in motivation, behavioral engagement, and academic performance between students who were in the control group and the badging experimental groups (participation-based or skills-based badge intervention)? (QT) Q2: Are differences in motivation, behavioral engagement, and academic performance moderated by gender? (QT) Q3: How do students in the experimental groups (participation-based badges, skill-based badges), perceive that the badging intervention influenced their motivation, behavioral engagement, and academic performance? (QL) Q4: How do students with different genders perceive that badging influenced their motivation, behavioral engagement, and academic performance? (QL) The hypothesis of this study is that gamification will have a positive effect on learning. More specifically, that badges have an impact on student motivation, behavioral engagement, and academic performance. Huang et al., (2020) in a meta-analysis of 30 gamification studies

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

confirm that gamification in educational studies has produced mixed results but overall, the gamification condition has produced a small to medium effect size, with g=.464 in favor of the gamification condition. Zhan et al., (2022) noted that gamification had the largest effect on the outcome of motivation. In findings from studies such as Ortega-Arranz et al., (2019), Rodrigues et al., (2022), and Sanchez et al., (2020) suggest that badges can promote behavioral engagement. Haruna et al., (2018) and Legaki et al., (2020) found that significant gains were made in academic performance in a gamified classroom. When examining the variable of gender, some studies have found no differences in gender in terms of retention (Sanchez et al., 2020) or motivation (Rincon-Flores et al., 2022), while others have found differences in gender in terms of types of motivation (Polo-Peña et al., 2021) and academic performance (Legaki et al., 2020; Rodrigues et al., 2022). Furthermore, the purpose of this study was to help understand more about implementing badging (participationbased and skill-based) and leaderboards in an accounting undergraduate course and if gender had any impact on results. Method This study used an embedded or nested, mixed-method design. Mixed methods approach allowed us to understand if the badging intervention had an impact on outcomes, as well as the process and experience of the badging intervention. Due to the constraints of university scheduling where students select their own sections, the paper relies on a quasi-experimental design rather than a true experiment, limiting full causal control. Participants and Context The participants included 202 students who were introductory undergraduate financial accounting students from business and accounting programs who were taking the first required accounting course in a traditional, face-to-face setting. The setting was an undergraduate university, in the western United States, with over 65% comprising of international students. The average age of students was 19. Three sections of the same course were taught in the same semester by the same instructor. IRB approval (protocol number: 2022-00434) was obtained before commencing the study. The researcher gained access to the participants during the fall of 2022 and spring semester of 2023. During the fifth week of the semester, the researcher distributed a survey called the Situational Motivation Scale (SIMS) to measure motivation. Additionally, the researcher asked students to provide demographic information on gender, cumulative GPA, and their preferences and experiences with video games as well as their perceptions about learning opportunities and video games. A non-binary gender variable was incorporated in the survey to be more inclusive as suggested by Cameron and Stinson (2019). During the intervention, the researcher collected LMS data analytics on participation, activity time, and page views. At the end of the intervention, the researcher administered the SIMS survey again.

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Qualitative data was collected through focus group discussions, that explored students’ perceptions of how the badging intervention influenced their motivation, behavioral engagement, and academic performance. Purposive sampling was used to ensure representation by gender for the focus groups (Onwuegbuzie & Collins, 2007). The researcher evaluated the qualitative information for saturation or until evidence of redundancy was achieved. Guest et al. (2006) posited that data saturation is possible with 12-15 participants, but more participants can be added if the researcher does not believe that they have reached saturation. After 10 focus group discussions, we reached saturation, taking the participant count to 30. Instrumentation and Procedures The SIMS instrument is a versatile self-report measure that asks a total of sixteen questions with four questions each on the following constructs: intrinsic motivation, integrated regulation, extrinsic regulation, and amotivation. The instrument uses a 7-point Likert scale to answer questions such as: Why are you currently engaged in optional quizzes? “Because I think that this activity is interesting” or “because I don’t have a choice” (Guay et al., 2001). The scores were averaged and reported. The assumption is that higher the intrinsic motivation and integrated regulation, the better the likelihood that students will have positive educational outcomes (Deci & Ryan, 2008). The behavioral engagement variable was captured by data analytics provided by Canvas, the LMS used by the university. Canvas offers log data according to student profile that tracks activity time, page views, and participation, that offers information about how students are participating and viewing the course. Additionally, the students’ Exam 2 score was also collected. The qualitative focus group protocol included questions such as “How would you describe your experience with badges during an introduction to accounting course? What was the most rewarding and most challenging part of your introduction to accounting course?” And “If you were giving advice to an instructor considering implementing badges and leaderboards in their courses, what would you say?” Treatment All three groups of the course had a leaderboard integrated into Canvas, but one served as a control group in terms of badging (no badges). Two groups had both the badging and leaderboard system integrated into their Canvas LMS learning course. The chosen software of BADGR was used for the study as it was the most aligned with the badge intentions of being an educational badge, awarded for completing tasks. Badges were awarded for completing optional quizzes for experimental groups. The first experimental group had badges awarded for effort or participation (e.g., completing the optional assignment). The second experimental group had skill-based badges (based on the score received on optional quizzes). Optional quizzes consisted of multiple-choice questions, fill-in-the blank questions, matching the term

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

to the definition questions, and categorization questions to simulate creating accounting journal entries. Optional quizzes were single-level (a task performed once with one level of difficulty) and scored to provide feedback on how proficient students were becoming on learning objectives similar to Barata et al., (2013). Students were not allowed to retake optional quizzes but were given the correct answer and feedback after completing the quizzes. Timing of Implementation The gamification intervention was implemented from week 5 to week 10 in the courses included in this study based on best practices in the literature. Kim and Castelli (2021) suggest that gamification interventions should be short-term to have the most successful implementation. Van Roy and Zaman (2018) found a decline (from the start of the gamification intervention) in intrinsic motivation at every 5-week interval of a fifteen-week course. Rodrigues et al. (2022) found a U-shape effect of motivation where the novelty wears off after about four weeks and then recovered in week 10 (familiarization effect) of a 14-week semester. Previous course performance (of the proposed introduction to financial accounting course) has indicated a drop in Exam scores from Exam 1 (at week 4.5) to Exam 2 (Week 10) by an average score of 8% (from a 77.5% average to a 69.5% average). This could possibly be a result of the mid-semester slump as indicated by Bolton (2003). Bolton (2003) noted that at the mid-point in the semester, energies ebb and students experience stagnation, and apathy. Additionally, Bolton (2003) calls the weeks before or after an exam or the weeks after the midterm “the doldrums” where students experience waning enthusiasm, a lack of motivation, or interest. Kyewski and Krämer (2018) used a five-week intervention time frame scheduled at the beginning of a fourteen-week course. Kyewski and Krämer (2018) found that student motivation significantly decreased from the beginning of the class to the end of the class. As such, the gamification intervention was implemented from week 5 to week 10 in the courses included in this study. Format of the Course and Badge Aesthetics The course covered a new chapter of content for four weeks and then the week before Exam 2 was a review week. Students were to complete compulsory homework assignments before every class period and for supplementary material, seven optional quizzes were offered. Optional quizzes included one vocabulary quiz and six quizzes covering approximately six learning objectives. A total of 28 optional quizzes with corresponding badges were offered to students. Badges were designed by the researcher using Accredible.com and were designed similar to Ortega-Arranz et al.’s (2019). See Table 1 for badge designs and earning criteria. BADGR software was used to distribute the badges but the leaderboard was updated manually - similar to the method used by Rincon-Flores et al. (2022). The researcher used an edit.org video game leaderboard design in designing the leaderboard (https://edit.org/edit/all/21jywx8ai). To draw attention to the leaderboard, students who entered the course in the LMS were directed to a

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

virtual trophy room that showed the leaderboard and attainable badges (for experimental groups). Table 1 Criteria for Awarding Badges Including Participation Badges and Skill-Based Badges Badge Name

Criteria Description

Participation badges Quiz Gold Badge

Completed quiz

Skill-based Badges Quiz Gold Badge (Chapter 6, 7, 8, 9)

90-100% of quiz answers correct

Quiz Silver badge

70-89% of quiz answers correct

Quiz bronze badge

50-69% of quiz answers correct

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Method of Data Analysis This study displays the mean and standard deviation of students’ responses to the SIMS instrument and inferential statistics Analysis of Variance (ANOVA) and Multivariate Analysis of Variance (MANOVA) with covariates or MANCOVA, to see changes (if any) between motivation at the beginning and end of the intervention. Guay et al., (2001) studied the SIMS instrument in five independent studies and found the majority of the studies had acceptable to high Cronbach Alpha’s when looking at the subscales of intrinsic motivation, identified regulation, external regulation and amotivation. For self-reported scale instruments, an acceptable range is .70–.80 for research purposes (Nunnally, 1978). For qualitative data analysis, all focus groups were voice recorded with consent of the participants. Voice recordings were transcribed and then checked for accuracy by the researcher. The verbatim transcript was sent to focus group participants for review/accuracy as a form of member-checking. The focus groups lasted 30–57 minutes depending on the group size. Data was explored to see if there were codes, categories, or themes that linked to the research questions and self-determination theory. The researcher used reflexivity and an audit trail by commenting in the MAXQDA software to shape research approaches and evidence as data was analyzed as well as to control bias (Ary et al., 2019). Findings This study is based on 202 students who were enrolled across six sections of an introductory accounting course for two semesters (Fall 2022 and Spring 2023). Of the 202 students, 177 students (88%) completed the survey at the beginning of the intervention (covariate data gathered) out of which 103 identified as female (58%) and 74 male (42%) students. Of these 177 students, 156 (77%) students completed the survey both at the beginning and the end of the intervention (motivation data gathered). The gender breakdown of these 156 students was: 92 (59%) female and 64 (41%) male students. Among the 30 focus group participants 13 were from the participation badge group and 17 from the skills-based badge group. The section that follows begins with which covariates were examined followed by the analyses of differences among groups in motivation, behavioral engagement and academic performance, followed by a description of gender differences. Differences Between Groups Descriptive Statistics for Covariates The following variables were used as a covariate based on factors that are associated with motivation in a gamified learning experience which include cumulative GPA, game learning opportunities, game experience, and game preferences. Covariates are used in the analysis to control for pre-existing achievement and game preferences. The mean cumulative GPA was similar for each of the groups ranging from (M = 3.53, SD = .44) in the control group to (M =

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

3.59, SD = .44) in the skills-based badges group. The lowest cumulative GPA reported across all groups was 1.30 and the highest GPA reported was a 4.0 GPA. Students were surveyed to understand game learning opportunities, game experience, and game preferences. Examples of questions asked included: video games offer opportunities to experiment with knowledge, I describe myself as a gamer, and I would vote in favor of using video games in the classroom. This study used scaled questions (ranging from 0-100) as a covariate to control for these experiences/perceptions. The questions were scaled on a 0-100 basis because covariates are traditionally measured on a continuous scale instead of a categorical scale (Pituch & Stevens, 2015). Selection of Covariates The selection and inclusion of covariates in the statistical procedures were based on previous studies examining their effects on gamified learning experiences. To understand the relationship between covariates and the outcome variables, a Pearson’s r correlation was conducted. Covariates that were significantly correlated (p < .05) to the outcome variable were used in subsequent analyses. For example, cumulative GPA and game preferences were not significantly related to motivation, and therefore these covariates were not included in the analyses. Game experience and game preferences were significantly correlated to pageviews (an outcome measure). Game experience was significantly correlated to activity time. Cumulative GPA and game preferences were significantly correlated to Exam 2 percentages. Motivation by Groups At the beginning of the intervention in week 5, students rated intrinsic motivation and integrated regulation at a high average across all sections with intrinsic motivation at (M = 4.48, SD = 1.61) and integrated regulation at (M = 5.78, SD = 1.16). Extrinsic motivation had an average result across all sections with a mean of 3.8 (SD = 1.58). Amotivation was rated low with a mean of 2.54 (SD = 1.39). A preliminary ANOVA was conducted to understand if there were differences between groups at week 5. All groups (Control, Participation badge, Skill-based badge group) had similar means, with no significant differences at that point in time. At the end of the intervention in Week 10, the average across all sections indicated very similar results. See Table 2. No significant differences were noted. To answer part of RQ1, no difference in motivation was noted between groups. Although motivation scores remained statistically similar across groups, students described varying motivational influences and experiences that were not fully captured by the quantitative measures. The gamification intervention evoked both positive and negative outcomes in terms of motivation with some students reporting high intrinsic motivation and integrated regulation while others reporting high extrinsic regulation, amotivation or a change of motivation from novelty to discouragement. These contrasting experiences may have offset one another when averaged across the group, contributing to non-significant quantitative

100

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

results for the gamification intervention. Five themes were extracted from focus groups including all four types of motivation and a change in motivation or a novelty effect. Table 2 Average Motivation Score by Group Motivation Construct

Intrinsic Motivation Control Participation Badge Skill-based Badge Total Change Integrated Regulation Control Participation Badge Skill-based Badge Total Change Extrinsic Regulation Control Participation Badge Skill-based Badge Total Change Amotivation Control Participation Badge Skill-based Badge Total Change

Week 5 M (SD)

Week 10 M (SD)

Change IM (Week 10–Week 5) M (SD)

4.64 (1.46) 4.27 (1.84) 4.54 (1.49) 4.48 (1.61)

4.86 (1.54) 4.42 (1.62) 4.50 (1.55) 4.58 (1.57)

0.22 (0.08) 0.15 (-0.22) -0.04 (0.06) 0.10 (-0.04)

5.85 (1.08) 5.63 (1.31) 5.86 (1.08) 5.78 (1.16)

5.74 (1.21) 5.50 (1.26) 5.89 (1.11) 5.71 (1.20)

-0.11 (0.13) -0.13 (-0.05) 0.03 (0.03) -0.07 (0.04)

3.93 (1.58) 3.76 (1.59) 3.71 (1.59) 3.80 (1.58)

3.52 (1.45) 3.64 (1.52) 3.41 (1.72) 3.52 (1.57)

-0.41 (-0.13) -0.12 (-0.07) -0.30 (0.13) -0.28 (-0.01)

2.52 (1.47) 2.57 (1.28) 2.50 (1.43) 2.53 (1.39)

2.46 (1.38) 2.80 (1.41) 2.58 (1.52) 2.62 (1.44)

-0.06 (-0.09) 0.23 (0.13) 0.08 (0.09) 0.09 (0.05)

Note. Sample size Control (n = 48), Participation Badge (n = 52) Skill-based Badge (n = 56)

Intrinsic Motivation (Interesting, Creative, Entertaining, Fun) Some participants from both groups (participation badge, skill-based badge) noted that the badges were interesting/creative and entertaining/fun. One student found badges more motivating than reading the textbook. Finally, one student commented on the surprise of seeing new elements of badges and leaderboards within the LMS of Canvas. Badges are like, really fun because it's different. Like I have not taken a class that use the same method of doing it and it just makes it more fun. It's kind of what you said like get a little game. I think playing it as a game is more motivating and attracting than just read the book (Participant 11, Female 6, Skill-based Badges)

101

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Integrated regulation (Enjoys Challenges and Aligns to Gamer Identity) A handful of students expressed that they loved challenges and even identified themselves as “a gamer” (Participant 7, Male 4, Skill-based Badges). This gaming identity seemed to encourage students to complete more optional quizzes. Students seemed to align their identities to the task which implies integrated regulation: “I love challenges. And so I kind of forced myself to do it” (Participant 5, Female 3, Participation Badges). Extrinsic Motivation (Recognition/Positive Feedback) Students commented that they enjoyed the recognition for their efforts via email. Further, the feedback email was more exciting and comforting instead of stressful feedback from a teacher. For example, Participant 4 said “I would say it got me motivated, to take to do the exercise. Because not only that, but like, knowing that there there's some recognition…I was like, Yeah, somebody saw my work and my effort” (Participant 4, Female 2, Participation Badges). Amotivation (Frustration) Both experimental groups reported frustration or amotivation, as they felt the leaderboard was more fictional, or not “real-life” and that it had no real impact on the course grade. Other students commented that they did not see the badge often and that it was a digital prize that was only made of pixels instead of tangible substance. “It's nothing I actually look at. Like, it's comparing that to like getting an actual trophy, I guess. You know, you show a trophy, but you don't show a digital prize because it's just pixels.” (Participant 14, Male 6, Participation Badges) Amotivation (Confusion with Intervention or Unfamiliarity) A handful of students noted that they felt there was a miscommunication of expectations with the badges and leaderboard. A couple of students could not find their pseudonym on the leaderboard, some had never experienced badges or leaderboards before, and a few were unsure of how the leaderboard was created. Change in Motivation (Novelty Effect then Discouragement) Some students, especially from the skills-based badge group, noted they enjoyed competition and challenging themselves, but they had fallen behind on the leaderboard and so they stopped doing the quizzes. Further, they started to focus on compulsory assignments or grades more than optional assignments. Overall, heterogenous findings were noted in terms of the motivation outcome.

102

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Behavioral Engagement by Groups The participation badge group had the highest number of quizzes attempted, activity time, and pageviews. Findings showed no significant differences between groups on activity time and pageviews, F(4, 350) = 1.795, p = .129, ηp² = .020. Because MANCOVA was not significant, no follow-up univariate ANOVA was necessary. Participations did not meet the normality assumption and so descriptive statistics were calculated. The percentage of the number of quizzes attempted by students, by group, was calculated with the maximum number of optional quizzes at 28. Over 80% of students in all groups attempted at least one optional quiz. Seven quizzes were created for each chapter. Around 60% of students completed seven or more quizzes, 38%-45% of students completed 14 or more quizzes and about 30% of students completed 21 or more quizzes. Students completed fewer quizzes as the weeks went on across all groups. Qualitative findings were mixed around behavioral engagement with students reporting within the two themes of either positive or no effects from the badges. Particularly female students commented more on the leaderboard aspect of the intervention, while male students tended to specifically mention or “hone in” on the badges. See Table 3 for behavioral engagement measures. Table 3 Behavioral Engagement Measures by Group1 Group

Control Total Participation Skills-based Total

56 62 59 177

Participations (Quiz Attempts) M (SD) 12.4 (10.4) 14.4 (10.5) 13.6 (10.6) 13.5 (10.5)

Activity time in hours (h) M (SD) 11.0h. (11.3h.) 11.5h. (12.9h.) 9.5h. (6.7h.) 10.7h. (10.6h.)

Pageviews M (SD) 613 (658) 717 (765) 606 (393) 647 (625)

Note. Maximum Participations (Quiz Attempts) was 28 quizzes

Learning Content/Chunking and Autonomy in Learning Some students reported that the badges helped them to prepare for class as they better understood what was being taught in class and it also helped with accounting equations and learning vocabulary, or accounting jargon. Others indicated that the badges helped them break course content into manageable chunks and track their progress throughout the semester. Additionally, several students reported that the badges fostered greater learning autonomy. Participant 2 (Male 2, Participation badges) noted that it is up to college students to take charge of their learning by realizing that although they are completing compulsory assignments for a grade, they may be lacking in their understanding of the content and so it is up to them to choose to complete optional assignments or seek supplemental materials. The same student

103

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

noted that there is freedom in optional assignments and that the quizzes could be a formative assessment to learn from. Did not Have Time to Participate or Did not see Value in Badges A large number of students lamented the time constraints of being a student. Although they might have wanted to participate in optional quizzes, they got off work late at night and only had a few hours to finish their assignments. They prioritized assignments that had course points over the optional assignments. Many students acknowledged the value of having the optional quizzes aligned to graded, in-class quizzes but did not see the value of the badges alone. Others noticed that badges were awarded through email, but they did not view their badges often. Supporting quotes were given in the “amotivation” theme listed above in motivation findings. Differences in Academic Performance Across Groups The control group had Exam 2 percentage scores of 64.68 (SD = 23.46). Participation badge group scores were 66.29 (SD = 22.46) and skill-based badge group scores were 70.00 (SD = 19.38). No significant differences were noted in the dataset between groups F (2, 174) = .656, p = .520. A small number of students remarked on the positive effect on academic performance outcomes. Qualitative findings included students explaining that optional quizzes were closely aligned to weekly, graded, in-class quizzes, but only a handful of students brought up the quizzes when preparing them for exam. Gender Differences in Motivation To understand if gender impacted outcomes, a two-way MANOVA was performed on all motivational constructs. No statistically significant results were noted, F (8, 294), 1.816, p = .074 but the two-way ANOVA of individual constructs indicated that Intrinsic Motivation (IM) and Amotivation (AM) had statistically significant findings. More investigation of the outcome showed a non-significant main effect of independent variables (group and gender), but a significant interaction between independent variables F (2, 153) = 3.419, p =.035, 𝜂!" = .044 For the change in Intrinsic Motivation (captured at the beginning and end of the intervention), a positive change would mean an increase in intrinsic motivation and a desirable result. See Table 4 for descriptive statistics. Figure 1 depicts the comparative graph for marginal means. The interaction effect (crossing lines) in Figure 1 occurs between the participation badge group and the control group and the control group versus the skill-based badge group. An inspection of the mean scores indicated that the participation badge group, males increased in intrinsic motivation (M = 0.39, SD = 0.94) and the skills-based badge group males also increased in motivation (M = 0.11, SD = 1.07) while motivation decreased in the control group males (M = -0.12, SD = 0.80). Alternatively, the control group females increased in intrinsic motivation (M =0.48, SD = 0.73) while experimental groups had smaller, less noticeable changes. This suggests that the combined influence of group and gender has a significant influence on intrinsic motivation within a gamified accounting course. At the end of the intervention, the males had higher intrinsic motivation in experimental groups compared to males in the control

104

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

group. For female students, the control group females had greater intrinsic motivation than experimental group females. Table 4 Intrinsic Motivation by Group and Gender Group and Gender

Week 5 M (SD)

Week 10 M (SD)

Control Total Female Male Participation Badge Total Female Male Skill-based Badge Total Female Male Grand Total

48 27 21 52 34 18 56 31 25 156

4.64 (1.46) 4.66 (1.69) 4.62 (1.15) 4.27 (1.84) 4.45 (1.83) 3.93 (1.87) 4.54 (1.49) 4.45 (1.64) 4.66 (1.30) 4.48 (1.61)

4.86 (1.54) 5.14 (1.65) 4.50 (1.33) 4.42 (1.62) 4.47 (1.59) 4.32 (1.72) 4.50 (1.55) 4.27 (1.76) 4.77 (1.24) 4.58 (1.57)

Change IM (Week 10–Week 5) M (SD) 0.22 (0.08) 0.48 (0.04) -0.12 (0.18) 0.15 (-0.22) 0.02 (-0.24) 0.39 (-0.15) -0.04 (0.06) -0.18 (0.12) 0.11 (-0.06) 0.10 (-0.04)

Note. Change is calculated by taking the mean at week 10 less the mean at week 5

Figure 1 Estimated Marginal Means of the Change in Intrinsic Motivation 1

Mixed results in Motivation with the Interaction of Groups and Genders Control group participants were not interviewed, and so no qualitative data was collected on this group. For the males in the experimental groups, one theory is that competition can bring out intrinsic motivation with immediate, relevant feedback and optimal challenges (Reeve & Ryan, 2021). With the challenges given, male students may have experienced task-orientation behaviors. Task orientation often called “task mastery” is aligned to intrinsic motivation or that individuals describe their efforts in learning as an improvement in an ability or skill (Jagacinski

105

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

& Strickland, 2000). The extrinsic motivators of the badge and leaderboard seemed to be supplemental or accessorial to the goal of learning, but still nice to see. Many male participants said that taking the optional quizzes was more about learning the content and making sense of the material. For example, Participant 1 said: It makes me feel that I accomplished the things because I can see my name on the leaderboard. Oh! I did it. There’s my name. But the studying is more like the main focus to do the optional quiz, but also the badges. It's not a main reason. (Participant 1, Male 1, Participation Badges) Participant 14 also mentioned: It's just extra practice because I won't get feedback on the actual homework until the next class period. I can get feedback on this immediately and then apply that to the homework (Participant 14, Male 6, Participation Badges) Additionally, some male participants noted that the quizzes gave a pattern or example to follow for required homework assignments and that allowed for more confidence in class and more confidence to do the assignment on their own. Another significant outcome was the change in amotivation for the interaction between group and gender. A decrease is a positive change as students do not feel helpless or have an apathy towards the activity/assignment. Males in the control group had a large decrease in their amotivation (M = -2.19, SD = 5.60) which is a good thing, implicating that students do not feel apathy toward the activity or perceive that the activity is useless. The results show F(2, 153) = 3.485, p =.033, ηp² = .044 compared to the participation badges male group that had a large increase in amotivation (M = 2.00, SD = 4.50). An increase in amotivation is a negative outcome in terms of motivation. See Table 5 and Figure 2. Table 5 Change in Amotivation by Group and Gender 11 n

Week 5 M (SD)

Week 10 M (SD)

48 27 21 52 34 18 56 31 25 156

2.52 (1.47) 1.92 (0.87) 3.30 (1.72) 2.57 (1.28) 2.51 (1.41) 2.67 (1.03) 2.50 (1.43) 2.17 (1.33) 2.92 (1.46) 2.53 (1.39)

2.46 (1.38) 2.23 (1.24) 2.75 (1.52) 2.80 (1.41) 2.60 (1.42) 3.17 (1.36) 2.58 (1.52) 2.18 (1.32) 3.09 (1.62) 2.62 (1.43)

Group and Gender Control Total Female Male Participation Badge Total Female Male Skill-based Badge Total Female Male Grand Total

106

Change AM (Week 10–Week 5) M (SD) -0.06 (-0.09) 0.31 (0.37) -0.55 (-0.20) 0.23 (0.13) 0.09 (0.01) 0.50 (0.33) 0.08 (0.09) 0.01 (-0.01) 0.17 (0.16) 0.09 (0.04)

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Figure 2 Estimated Marginal Means of Change for Amotivation 1

This intervention gave students participation badges based on completing optional quizzes. Deci et al., (2001) caution against giving completion-contingent rewards as these rewards can undermine intrinsic motivation. The reward itself carries little to no competence confirmation and so students are not likely to perceive competence. Without the perceived competence of knowing the accounting content and doing well on in-class quizzes, students can feel the negative effects of instructor control that the reward brings which might increase amotivation (Deci et al., 2001). A handful of males, from the participation badges focus group noted that the leaderboard was not a true representation of content knowledge as the number #1 spot was given to all students after completing the optional quizzes without any level of competency. Additionally, students noted that even though they did not get the score they wanted on optional quizzes, students were still rewarded with the leaderboard placement. Some indications of “ego orientation” were expressed. Ego orientation is when students assess their performance based on the rest of the class, which creates social comparison and the desire to win (Jagacinski & Strickland, 2000). This desire includes the feeling of being better than other participants and is often associated with negative behaviors such as lack of effort, persistence, and performance (Jagacinski & Strickland, 2000). I think the primary purpose of a leaderboard is meant to compare you to other people in your class. And I don't want to say like, oh, that sounds negative, I actually think that's a good thing to compare yourself to other people. (Participant14, Male 6, Participation badges). I don't really see a large benefits for students except for being able to do quantitate themselves in a sense of where do I stand with the rest of my class. But like I said, in its current system, that's really not possible…It just gives you the completion. So, everyone who did the quiz for that chapter, everyone's in first…There’s just no point (Participant 15, Male 7, Participation badges).

107

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Overall, no significant differences were found between the students’ beginning and end of the interventions’ Situational Motivation Scale when examining them by group, but when disaggregating the data by group and by gender, differences were displayed with the interaction between group and gender on the constructs of intrinsic motivation and amotivation. Behavioral Engagement and Academic Performance by Gender The variable of participations was disaggregated to number of quizzes attempted out of twentyeight quizzes. Furthermore, behavioral engagement and academic performance measures including activity time, pageviews and Exam 2 percentage were examined. See Table 6. Females showed significantly higher behavioral engagement in terms of pageviews F (2, 174) = 4.10, p = .044, ηp² = .023. Female pageviews were in the range of 105-3,981 pageviews while males had a range of 44-1,194 across the groups. Table 6 Behavioral Engagement Descriptive Statistics by Group and Gender Group n Quiz Activity time in Pageviews M (SD) hours (h) M (SD) M (SD) Control Total 56 12.4 (10.4) 11.0h. (11.3h.) 613 (658) Female 31 14.0 (10.0) 12.9h. (13.9h.) 764 (770) Male 25 10.4 (10.6) 8.7h. (6.9h.) 430 (437) Participation 62 14.4 (10.5) 11.5h. (12.9h.) 717 (765) Female 42 16.5 (10.1) 13.6h. (15.1h.) 780 (769) Male 23 11.5 (10.5) 8.7h. (8.3h.) 626 (763) Skills-based 59 13.6 (10.6) 9.5h. (6.7h.) 606 (393) Female 31 14.1 (11.0) 9.8h. (6.1h.) 696 (420) Male 28 13.1 (10.2) 9.5h. (7.3h.) 511 (343) Grand Total 177 13.5 (10.5) 9.5h. (6.7h.) 647 (625)

Exam 2 % M (SD) 66.6 (21.5) 65.9 (20.9) 67.3 (22.7) 67.4 (20.9) 69.7 (19.6) 63.3 (23.0) 70.9 (17.1) 69.9 (18.1) 72.0 (16.2) 66.6 (21.9)

Note. Maximum Participations (Quiz Attempts) was 28 quizzes

Some females in the participation badge group and skill-based badge group described a clicking behavior without putting in excessive effort to solve the problem. This seems to support the quantitative finding that female pageviews were significantly higher compared to their male counterparts. “If it's no pressure at all, you just pick something you don't really learn at all. It’s just for the sake of taking it and just seeing the answer after” (Participant 28, Female 12, Skill-based Badges). A large number of female participants in the focus group reported that they knew that the quizzes were optional and that the quizzes would not affect their grade. No significant differences were noted in Exam 2 percentages between groups, genders, or the interaction (group x gender). Overall, female participants tended to interact more with the LMS in terms of pageviews or clicking on the content. Descriptive statistics highlighted that the

108

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

clicking behavior seems congruent with the other behavioral engagement measures of females having more activity time and more quizzes attempted than males. Discussion Overall, findings revealed a disconnect between students' reported motivation and their actual quiz-taking behaviors. Participants further described badges as having limited value as rewards but identified the feedback and recognition they received through badging as beneficial outcomes. Furthermore, gamification experiences varied among student groups, and the implications of these differences are discussed. Negative and Positive Outcomes of Badging A primary finding of this study was the disconnect between students' reported motivation and their actual participation in supplemental quizzes. While students self-reported high initial motivation, their actual participation dropped sharply over time. This indicates that self-report inventory questionnaires may not fully capture authentic behavioral intentions. Meyer et al. (2018) found that students often overestimate their contribution to a class. It is important to note that a significant portion of students could not participate due to real-life constraints (e.g., working late) or chose to prioritize graded coursework over optional gamified elements. The design could be improved by better integrating gamification directly into core compulsory components or offering highly flexible, individualized engagement options. The Impact of Badges Depends on Their Perceived Value Additionally, students expressed frustration because the digital badges felt like "just pixels" and lacked tangible or grade-based substance. To improve the intervention, the rewards need to hold more concrete value or credibility for the students. This finding is congruent with Almeida et al. (2023) that online badges lacked credibility compared to a more tangible badge. More importantly, these findings suggest that the effectiveness of badges depends not merely on their presence but on the value students assign to them. From a Self-Determination Theory perspective, external rewards that are not personally meaningful are unlikely to strengthen motivation or influence behavior. In-game prizes and tokens must hold meaning to the student or motivation around the gamification may be low (Abadi et al., 2022). Some students also linked the quizzes to autonomy, describing participation as their own responsibility, consistent with Self-Determination Theory's autonomy component and prior work on self-directed learning paths (Rodrigues et al., 2022; Sasmaz et al., 2025). Recognition and Feedback May Be More Influential Than Rewards Although some participants questioned the value of badges as rewards, others reported appreciating the recognition and feedback associated with earning them. This finding suggests that the motivational benefits of badging may stem less from the reward itself and more from the information communicated through the badge. In other words, badges may function

109

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

primarily as indicators of progress, achievement, and instructor acknowledgment rather than as incentives designed to alter behavior. The appreciation students expressed toward receiving recognition is particularly noteworthy because it highlights the importance of competencerelated feedback. Within Self-Determination Theory, perceptions of competence are closely linked to motivation and engagement. Receiving acknowledgment for effort and accomplishment may reinforce students' awareness of their progress and contribute positively to their learning experience. This interpretation is consistent with Sasmaz et al. (2025), who emphasized the role of timely and meaningful feedback in supporting student learning and engagement. Gamification Experiences Differ Across Student Groups The findings also indicate that students do not experience gamification in the same way. Gender differences emerged in both motivation and behavioral patterns, suggesting that competitive and achievement-oriented elements may be interpreted differently across learners. Female students more frequently described the leaderboard as exciting and motivating and reported positive feelings associated with competition and recognition. In contrast, male students expressed more varied reactions, including both positive social comparison and negative emotions such as embarrassment or shame. These findings are consistent with Carpenter et al. (2018), who identified gender differences in motivational orientations toward competition. More broadly, the results suggest that students may differ in the extent to which they are motivated by ego-oriented goals, such as outperforming others, versus task-oriented goals, such as personal growth and mastery. Such differences highlight the limitations of a one-size-fits-all approach to gamification. Implications for Practice Collectively, these findings suggest that badges should not be viewed as a standalone mechanism for increasing student participation. Instead, their effectiveness appears to depend on the perceived value of the reward, the quality of feedback and recognition provided, and the extent to which the intervention accommodates individual differences among learners. Offering options such as competitive leaderboards, individual achievement badges, collaborative challenges, or noncompetitive recognition systems may allow students to select experiences that align with their preferences and motivational orientations. This flexible approach may create a more inclusive learning environment while maximizing the potential benefits of gamification across diverse student populations. Limitations This research study was not without limitations. Although cumulative GPA and other game preferences were controlled for by using a two-way MANCOVA analysis, quasi-experimental designs do not provide full control over the study as students are not randomly assigned to treatment groups (Ary et al., 2019). The study included a high proportion of international

110

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

students which may limit the transferability of results to other student populations. The intervention period was relatively short and is insufficient to capture the long-term effects of badging over time. Additionally, student motivations may be solely driven by academic performance and may not be fully captured by self-report inventory questionnaires. Likewise LMS analytics such as pageviews and activity time were used as proxies for behavioral engagement which may not accurately reflect meaningful learning engagement. Furthermore, students received extra credit incentives to fill out questionnaires, which could lead to selection bias and an unwillingness to give candid answers, while giving the most socially desirable responses (Ary et al., 2019). An additional research limitation included potential researcher bias in terms of reliability, as the researcher was both the co-investigator and instructor of the gamified course (Zohrabi, 2013). Future studies involving different student populations, independent researchers, and longer intervention periods could provide a more comprehensive understanding of the role of badging in supporting student motivation, behavioral engagement, and academic performance. Conclusion This study contributes to the literature by isolating one gamification element (badging) while holding the leaderboard fixed. To partial out pre-existing factors, cumulative GPA and game preferences were used as covariates. Limited research has been done on gamification in undergraduate accounting education. Findings indicate that badging may not be as impactful on motivation, behavioral engagement, and academic performance of students in an introductory accounting course. Positive outcomes of feedback and recognition were reported but students reported that badges were not meaningful rewards. Although students expressed excitement about the intervention, this did not result in long-term, optional quiz taking behaviors. In examining gender differences, females and males experienced the gamification differently in terms of intrinsic motivation and amotivation across groups. Additionally, females had more pageviews than male participants. No differences by gender were noted in terms of academic performance. Insight gained can help instructors assess the best practices of gamification implementation with regards to motivation, behavioral engagement, and academic performance in higher education.

Declaration of Generative AI and AI-Assisted Technologies in Writing Process AI tools were used in the preparation of this manuscript to help with revising for conciseness. Additionally, AI was used to produce sample badge designs.

111

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

References Abadi, B. M. F., Samani, N. K., Akhlaghi, A., & Najibi, S. (2022). Pros and cons of tomorrow’s learning: A review of literature of gamification in education context. Medical Education Bulletin, 3(4), 543–554. https://doi.org/10.22034/meb.2022.350941.1063 Almeida, C., Kalinowski, M., Uchôa, A., & Feijó, B. (2023). Negative effects of gamification in education software: Systematic mapping and practitioner perceptions. Information and Software Technology, 156, Article 107142. https://doi.org/10.1016/j.infsof.2022.107142 Alvi, I. (2025). Language learners’ intentions to use gamified learning apps: The role of gameful experience and cognitive engagement. Journal of Education: Technology in Education, 13(2), 87–112. Ary, D., Jacobs, L. C., Sorensen, C. K., & Walker, D. A. (2019). Introduction to research in education (10th ed.). Cengage. Barata, G., Gama, S., Jorge, J., & Gonçalves, D. (2013). Improving participation and learning with gamification. In Proceedings of the First International Conference on Gameful Design, Research, and Applications (pp. 10–17). https://doi.org/10.1145/2583008.2583010 Bolton, M. J. (2003). Overcoming inertia: Guiding criminal justice students through midsemester slump. Journal of Criminal Justice Education, 14(2), 355–370. Cainas, J. M., & Kralik, J. M. (2024). The great accounting escape: A teaching tool for relevant costing and short-term decisions. Issues in Accounting Education, 39(1), 123–133. https://doi.org/10.2308/ISSUES-2022-003 Cameron, J. J., & Stinson, D. A. (2019). Gender (mis)measurement: Guidelines for respecting gender diversity in psychological research. Social and Personality Psychology Compass, 13(11). https://doi.org/10.1111/spc3.12506 Carpenter, J., Frank, R., & Huet-Vaughn, E. (2018). Gender differences in interpersonal and intrapersonal competitive behavior. Journal of Behavioral and Experimental Economics, 77, 170–176. https://doi.org/10.1016/j.socec.2018.10.003 Chugh, R., & Turnbull, D. (2023). Gamification in education: A citation network analysis using CitNetExplorer. Contemporary Educational Technology, 15(2). https://doi.org/10.30935/cedtech/12863 Deci, E. L., Eghrari, H., Patrick, B. C., & Leone, D. R. (1994). Facilitating internalization: The self-determination theory perspective. Journal of Personality, 62(1), 119–142. https://doi.org/10.1111/j.1467-6494.1994.tb00797.x Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. Plenum Press. Deci, E. L., & Ryan, R. M. (1994). Promoting self-determined education. Scandinavian Journal of Educational Research, 38(1), 3–14. https://doi.org/10.1080/0031383940380101 Deci, E. L., Koestner, R., & Ryan, R. M. (2001). Extrinsic rewards and intrinsic motivation in education: Reconsidered once again. Review of Educational Research, 71(1), 1–27. https://doi.org/10.3102/00346543071001001

112

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Deci, E. L., & Ryan, R. M. (2008). Facilitating optimal motivation and psychological wellbeing across life’s domains. Canadian Psychology/Psychologie Canadienne, 49(1), 14–23. https://doi.org/10.1037/0708-5591.49.1.14 Dichev, C., & Dicheva, D. (2017). Gamifying education: What is known, what is believed and what remains uncertain: A critical review. International Journal of Educational Technology in Higher Education, 14(9). https://doi.org/10.1186/s41239-017-0042-5 Fredricks, J. A., Blumenfeld, P. C., & Paris, A. H. (2004). School engagement: Potential of the concept, state of the evidence. Review of Educational Research, 74(1), 59–109. https://doi.org/10.3102/00346543074001059 Goulding, J., Sharp, H., & Twining, P. (2024). Awarding digital badges: Research from a first-year university course. Higher Education Research & Development, 43(3), 640– 656. https://doi.org/10.1080/07294360.2024.2315039 Guay, F., Vallerand, R. J., & Blanchard, C. (2001). On the assessment of situational intrinsic and extrinsic motivation: The Situational Motivation Scale (SIMS). Motivation and Emotion, 25, 179–213. Guest, G., Bunce, A., & Johnson, L. (2006). How many interviews are enough? An experiment with data saturation and variability. Field Methods, 18(1), 59–82. https://doi.org/10.1177/1525822X05279903 Haruna, H., Hu, X., Chu, S. K. W., Mellecker, R. R., Gabriel, G., & Ndekao, P. S. (2018). Improving sexual health education programs for adolescent students through gamebased learning and gamification. International Journal of Environmental Research and Public Health, 15(9), Article 2027. https://doi.org/10.3390/ijerph15092027 Huang, R., Ritzhaupt, A. D., Sommer, M., Zhu, J., Stephen, A., Valle, N., Hampton, J., & Li, J. (2020). The impact of gamification in educational settings on student learning outcomes: A meta-analysis. Educational Technology Research and Development, 68(4), 1875–1901. https://doi.org/10.1007/s11423-020-09807-z Jagacinski, C. M., & Strickland, O. J. (2000). Task and ego orientation: The role of goal orientations in anticipated affective reactions to achievement outcomes. Learning and Individual Differences, 12(2), 189–208. https://doi.org/10.1016/S10416080(01)00037-1 Kim, J., & Castelli, D. M. (2021). Effects of gamification on behavioral change in education: A meta-analysis. International Journal of Environmental Research and Public Health, 18(7), Article 3550. https://doi.org/10.3390/ijerph18073550 Kocsis, Á., & Molnar, G. (2025). Factors influencing academic performance and dropout rates in higher education. Oxford Review of Education, 51(3), 414–432. https://doi.org/10.1080/03054985.2024.2316616 Kyewski, E., & Krämer, N. C. (2018). To gamify or not to gamify? An experimental field study of the influence of badges on motivation, activity, and performance in an online learning course. Computers & Education, 118, 25–37. https://doi.org/10.1016/j.compedu.2017.11.006 Laranjeira, M., & Teixeira, M. O. (2025). Relationships between engagement, achievement and well-being: Validation of the engagement in higher education scale. Studies in Higher Education, 50(4), 756–770. https://doi.org/10.1080/03075079.2024.2354903

113

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Legaki, N. Z., Xi, N., Hamari, J., Karpouzis, K., & Assimakopoulos, V. (2020). The effect of challenge-based gamification on learning: An experiment in the context of statistics education. International Journal of Human–Computer Studies, 144, Article 102496. https://doi.org/10.1016/j.ijhcs.2020.102496 Meyer, M. L., McDonald, S. A., DellaPietra, L., Wiechnik, M., & Dasch-Yee, K. (2018). Do students overestimate their contribution to class? Congruence of student and professor ratings of class participation. Journal of the Scholarship of Teaching and Learning, 18(3). https://doi.org/10.14434/josotl.v18i3.21516 Nunnally, J. C. (1978). Psychometric theory (2nd ed.). Jossey-Bass. Oliveira, W., Hamari, J., Shi, L., Toda, A. M., Rodrigues, L., Palomino, P. T., & Isotani, S. (2023). Tailored gamification in education: A literature review and future agenda. Education and Information Technologies, 28(1), 373–406. https://doi.org/10.1007/s10639-022-11122-4 Onwuegbuzie, A. J., & Collins, K. M. T. (2007). A typology of mixed methods sampling designs in social science research. The Qualitative Report. https://doi.org/10.46743/2160-3715/2007.1638 Ortega-Arranz, A., Er, E., Martínez-Mones, A., Bote-Lorenzo, M. L., Asensio-Pérez, J. I., & Muñoz-Cristóbal, J. A. (2019). Understanding student behavior and perceptions toward earning badges in a gamified MOOC. Universal Access in the Information Society, 18(3), 533–549. https://doi.org/10.1007/s10209-019-00677-8 Pituch, K. A., & Stevens, J. P. (2015). Applied multivariate statistics for the social sciences (6th ed.). Routledge. Polo-Peña, A. I., Frías-Jamilena, D. M., & Fernández-Ruano, M. L. (2021). Influence of gamification on perceived self-efficacy: Gender and age moderator effect. International Journal of Sports Marketing and Sponsorship, 22(3), 453–476. https://doi.org/10.1108/IJSMS-02-2020-0020 Reeve, J., & Ryan, R. M. (2021). Intrinsic motivation, psychological needs, and competition: A self-determination theory analysis. In The Oxford Handbook of the Psychology of Competition. Oxford University Press. https://doi.org/10.1093/oxfordhb/9780190060800.013.10 Rincon-Flores, E. G., Mena, J., & López-Camacho, E. (2022). Gamification as a teaching method to improve performance and motivation in tertiary education during COVID19. Education Sciences, 12(1), Article 49. https://doi.org/10.3390/educsci12010049 Rodrigues, L., Pereira, F. D., Toda, A. M., Palomino, P. T., Pessoa, M., Carvalho, L. S. G., Fernandes, D., Oliveira, E. H. T., Cristea, A. I., & Isotani, S. (2022). Gamification suffers from the novelty effect but benefits from the familiarization effect: Findings from a longitudinal study. International Journal of Educational Technology in Higher Education, 19(1). https://doi.org/10.1186/s41239-021-00314-6 Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68– 78. Sanchez, D. R., Langer, M., & Kaur, R. (2020). Gamification in the classroom: Examining the impact of gamified quizzes on student learning. Computers & Education, 144, Article 103666. https://doi.org/10.1016/j.compedu.2019.103666

114

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Sasmaz, M. B., Frost, R. D., Gordon, E. R., Kenyo, L. N., Matta, V. A., & Raisch, M. E. (2025). Scoreboard for Excel: The design and implementation of software to advance active learning, automate grading, and reduce cheating. Issues in Accounting Education, 1–21. https://doi.org/10.2308/ISSUES-2023-044 Steenkamp, N., Fisher, R., & Nesbit, T. (2024). Understanding accounting students’ intentions to use digital badges to showcase employability skills. Accounting Education, 33(6), 906–934. https://doi.org/10.1080/09639284.2023.2276200 Tahir, F., Mitrovic, A., & Sotardi, V. (2022). Investigating the causal relationships between badges and learning outcomes in SQL-Tutor. Research and Practice in Technology Enhanced Learning, 17(7). https://doi.org/10.1186/s41039-022-00180-4 Velazquez-Garcia, L., Longar-Blanco, M. D. P., Cedillo-Hernández, A., & Bustos-Farías, E. (2024). Gamification in the classroom: Motivating higher education students using digital badges. International Journal of Learning and Teaching, 10(4), 532–538. https://doi.org/10.18178/ijlt.10.4.532-538 Van Roy, R., & Zaman, B. (2018). Need-supporting gamification in education: An assessment of motivational effects over time. Computers & Education, 127, 283–297. https://doi.org/10.1016/j.compedu.2018.08.018 Veiga, F., Wong, Z. Y., Veiga, F. M., Seabra, F., Festas, I., Faria, L., Forno, L. F., & Martínez, I. (2026). Higher education student engagement in the academic community: Clarifying the concept and proposing a short measure. Frontiers in Education, 11, Article 1773320. https://doi.org/10.3389/feduc.2026.177332 Zhan, Z., He, L., Tong, Y., Liang, X., Guo, S., & Lan, X. (2022). The effectiveness of gamification in programming education: Evidence from a meta-analysis. Computers and Education: Artificial Intelligence, 3, Article 100096. https://doi.org/10.1016/j.caeai.2022.100096 Zohrabi, M. (2013). Mixed method research: Instruments, validity, reliability and reporting findings. Theory and Practice in Language Studies, 3(2), 254–262. https://doi.org/10.4304/tpls.3.2.254-262

Corresponding author: Tialei Scanlan Email: tialei.scanlan@byuh.edu

115

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

116

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Relationship Between School-Related and Private Digital Media Use and Self-Learning Competencies Among German Vocational Students Maxi Eileen Brausch-Böger Technical University of Munich, Germany Manuel Förster Technical University of Munich, Germany

117

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Abstract Digital media has gained significant relevance in vocational education and training (VET) in recent years, in educational policy as much as in day-to-day classroom practice. However, little is known about how students’ use of digital media relates to their self-learning competences (SLC). This study examines this relationship among German vocational school students (n = 248) using a quantitative cross-sectional design. SLC was measured with Arnold’s Competence Questionnaire, while digital media use was assessed using the framework of the International Computer and Information Literacy Study. Structural Equation Modeling (SEM) was used to analyze the data. We analyzed four dimensions of digital media use: use for lesson preparation, use of software programs, use for private purposes, and use for other school-related purposes, along with five dimensions of SLC: professional, methodical, personal, emotional, and social. Results indicated school-related digital media use was positively associated with methodical and personal competence. SEM additionally indicated positive associations with emotional and social competence. In contrast, use for private purposes was negatively associated with methodical competence, and no significant relationship emerged between digital media use and professional competence. Overall, the findings suggest that the way digital media are integrated into learning processes is more closely linked to SLC than media use per se. Purposeful, learning-oriented use was related to several competence dimensions, whereas unspecific use showed no such associations. The findings have implications for teaching practice, teacher education, and future research. Keywords: digital media, self-directed learning, self-learning competence, vocational education and training, structural equation modeling, technology in education

118

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

The integration of digital media such as learning management systems, mobile devices, digital learning platforms, and online communication tools has changed teaching and learning processes in vocational education and training (VET). Digital media is an important part of modern educational practice (Alam et al., 2025; Schulte et al., 2014). Technological developments connected to Industry 4.0 are changing skill requirements in the labor market and increasing pressure on educational institutions to adapt existing teaching approaches (Backes-Gellner & Lehnert, 2023; Bläsche, 2018; Findeisen & Wild, 2022). Here, VET is expected not only to teach occupation-specific knowledge but also to support lifelong learning and digital and autonomous learning competences (Müller & Nannen-Gethmann, 2020). Vocational schools also play an important role in digital education, although challenges such as limited technological expertise among teachers remain (Bläsche, 2018; Hannack & Wucherpfennig, 2018). Overall, these developments increase the importance of learner autonomy and more flexible forms of learning (Eickelmann, 2017). Digital transformation is also changing the way learning is understood and organized. Digital technologies create new ways to learn, communicate, and collaborate enabling more personalized learning (Pilz & Fürstenau, 2019). In this context, the ability to learn independently becomes more important. In the literature, this is referred to as self-directed learning, which is learners’ ability to plan, implement, and assess their own learning processes (Knowles, 1975). Although learner-centered approaches are gaining more attention, their practical application is still a challenge, especially in structured educational settings like VET (European Centre for the Development of Vocational Training [CEDEFOP], 2022). An essential prerequisite of successful learning is self-learning competence (SLC), the capacity to learn independently, as well as to organize and apply knowledge, skills, and attitudes to learning (Arnold et al., 2003). It is cognitive, motivational, and emotional in nature and contributes to self-directed learning (Arnold et al., 2003). These competences are most apparent in digital learning settings in which learning is done more independently (Beyer, 2020). With digital media, competence development can be fostered, for example in collaborative learning spaces where communication and cooperation are enhanced (Seever & Schacher, 2022). But the extent to which students benefit from these opportunities depends on their ability to regulate their own learning which is closely linked to the development of SLC (Zumbach & Astleitner, 2016). While these opportunities exist, integrating digital media into VET is not without challenges. Moreover, studies indicate adverse effects of digital media on students’ well-being and mixed findings on its impact on learning processes (Kovalchuk & Sheludko, 2019; Tække & Paulsen, 2021). These findings demonstrate the importance of pedagogically grounded approaches that balance digital media use with promoting autonomous learning. The term digital media is used for any kind of media encoded in a machine-readable format such as text, audio, video and graphical materials (Bremer & Antony, 2013; Eickelmann, 2017; Eickelmann et al., 2019; Fraillon et al., 2014). These can be used in many ways in educational settings and informal learning environments (Eickelmann & Gerick, 2017; Hugo et al., 2022). Meta-analysis of SRL in vocationally oriented learning contexts indicates that learners’ use of self-regulated learning strategies is related to learning performance and learner engagement

119

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

(Hemmler & Ifenthaler, 2024). While self-directed learning is widely recognized as an important factor in the use of digital media for learning purposes (Beyer, 2020; Jossberger et al., 2020), there is still very limited empirical research related to the use of digital media in relation to specific aspects of SLC in VET. Against this background, the present study aims to contribute to a better understanding of the relationship between digital media use and SLC among vocational students in Germany. The study examines students’ current level of SLC and their use of digital media for learning purposes. In addition, the study explores whether different forms of digital media use are related to specific dimensions of SLC. Although digital media plays an important role in VET, little is still known about how these forms of media use relate to different aspects of SLC. The findings may help to better understand how learning environments can be designed and how teacher education and educational policy in vocational education and training can be further developed. Theoretical Framework Self-Learning Competence (SLC) The concept of self-learning competence (SLC) is grounded in the broader framework of selfdirected learning (SDL) and self-regulated learning (SRL). Self-directed learning is a dynamic learning process in which individual responsibility is taken for learning needs, goals, strategies, and the process of monitoring and adapting (e.g., Knowles, 1975; Schiefele & Pekrun, 1996; Weinert, 2000). While self-directed learning primarily refers to the learning process itself, SLC focuses more on the underlying competences that allow such learning to take place (Arnold & Gómez Tutor, 2006). Self-regulated learning, in turn, describes learners as active individuals who plan, monitor, and evaluate their own learning processes and regulate the motivational, emotional, and social conditions of learning (Zimmerman, 2000, 2002). Within this perspective, the SLC construct according to Arnold et al. (2003) offers a domain-specific operationalization for vocational education and training that differentiates the competences underlying self-regulated learning. According to Arnold et al., SLC comprises six interrelated dimensions: methodical, personal, emotional, professional, social, and communication competence (see Figure 1). In this sense, SLC is an important prerequisite for self-directed and self-regulated learning, equipping learners with the competences needed to engage in autonomous, goal-oriented learning activities. The SLC dimensions correspond to central components of SRL. First, methodical competence reflects the metacognitive strategies of planning, monitoring, and reflection. Second, personal competence captures motivational aspects such as self-efficacy and perseverance. Third, emotional competence relates to the regulation of emotions in learning contexts. Fourth, social competence corresponds to co-regulatory processes such as help-seeking and collaborative learning (Panadero, 2017; Zimmerman, 2000). Professional competence, in turn, represents the domain-specific knowledge base to which these regulatory processes are applied. Arnold et al.’s (2003) framework was used as the basis for measurement because it was developed

120

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

specifically for the German VET context and provides validated items for the target group of this study. Of the six dimensions, communication competence, which is defined as the ability to purposefully shape cooperative and communication processes in an effective and conflict-free manner (Arnold et al., 2003), was not included in the present study for two reasons. First, this dimension refers to the active, goal-directed shaping of interaction processes, which could not be adequately captured within the self-report format and scope of the present survey. Second, the interactive aspects most relevant to self-regulated learning, such as exchanging ideas with classmates and engaging in discussions about learning content, are partially reflected in the social competence dimension. The present study therefore focuses on the five remaining dimensions. Figure 1 Model of Self-learning Competence adapted from Arnold et al. (2003)

Professional Competence Professional competence describes an individual’s knowledge or skills regarding a particular subject, which consists, for example, of previous knowledge or existing general knowledge. This prior knowledge is necessary to ensure a successful self-learning process. Therefore, in didactic settings, it is essential to enable learners to connect their previous experiences to their current learning (Arnold et al., 2003).

121

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Methodical Competence Methodical competence refers to the ability to organize, reflect on, and shape one’s own learning processes. It includes the use of suitable learning and problem-solving strategies as well as metacognitive elements such as planning, monitoring, and evaluating learning activities (Arnold et al., 2003). In the context of self-directed learning, methodical competence helps learners structure and manage their learning processes independently. Personal Competence Personal competence describes the ability to take responsibility for one’s own actions and learning processes. It includes aspects such as self-confidence, motivation, perseverance, and the development of a stable sense of identity. These competencies support learners in independently managing and maintaining their learning activities over time. According to Arnold et al. (2003), experiences of autonomy, competence, and personal engagement are central indicators of personal competence, which is closely related to self-determination theory (Deci & Ryan, 1993). Emotional Competence Emotional competence is an interplay of self-perception, positive self-esteem, and social awareness, such as empathy. It is also described as an awareness of one’s emotions and effects. In doing so, learners become aware of their strengths and weaknesses. Likewise, the individual can self-regulate their emotions, such as by controlling the expression of feelings (Arnold et al., 2003). Social Competence Social competence refers to the learners’ ability to cope with conflicts, make contact, and work in teams. It involves seeking out exchanges of experience and learning with like-minded individuals to develop new plans and goals from these interactions (Arnold et al., 2003). In vocational education, SLCs are crucial because they enable students to take control of their learning journey, fostering autonomy and flexibility in a fast-paced and dynamic work environment. These skills can be cultivated through guided learning (Jossberger et al., 2020), tailored learning environments (Mejeh & Held, 2022), and workplace learning (Pylväs et al., 2022). Digital tools and resources provide learning opportunities online, including tutorials and communities that complement formal education. Digital media can also facilitate self-directed learning as students may need to organize and manage parts of their learning process in a new work environment.

122

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Role of Digital Media in VET Digital media integration into VET is increasingly important in today’s labour market context. Digitalization and increasing relevance of academic qualifications are also shaping the dual system (Haasler, 2020) which combines workplace training with classroom learning. Digital technologies (smartphones, tablets, smart boards, etc.) have become increasingly common in education (Tække & Paulsen, 2021). These technologies often indicate new opportunities for learning (UNESCO, 2024). Digital media has also changed the way we learn. Students can access information and communicate with others inside and outside the classroom. So learning is no longer just tied to fixed classroom conditions and can happen in more flexible and continuous ways (Tække & Paulsen, 2021). Eickelmann et al. (2019) categorize digital media use in education into three distinct categories: in-classroom, out-of-classroom and non-school related. In-classroom use involves tools like computers for assignments and lessons. Students increasingly rely on digital media for homework and research outside the classroom, indicating a shift towards home-based learning in places like Germany (Eickelmann, 2019). Non-school-related use, such as social media or gaming, presents both benefits and challenges. Digital media can support collaborative learning and provide creative expression. At the same time, it may also distract students and influence learning differently depending on whether it is used in academic or private contexts (Eickelmann et al., 2019). SLC and Digital Media in VET In this study, we distinguish between two overarching categories of digital media use. Schoolrelated use refers to the use of digital media initiated by or directed toward vocational schooling, such as preparing for lessons, using learning-related software, or completing other school-related tasks. Private use, by contrast, refers to media use in learners’ everyday lives outside of school requirements, such as social media, online communication, and entertainment. Workplace-based media use is not considered in this study. School-related Use of Digital Media Digital media has become an important aspect of vocational education and offers new possibilities for interactive and practice-oriented learning (Cattaneo et al., 2021; Ramin, 2022; Schulte et al., 2014). Digital tools in teaching can enhance the professional competencies of students and prepare them for digital workplaces (Cattaneo et al., 2021; Schulte et al., 2014). Digital media can also support learning environments, with respect to active participation, feedback, and problem-based learning (Gräsel et al., 2020). E-learning resources and digital tools may enhance students’ ability to learn independently by promoting the planning and regulation of learning processes (Barz et al., 2024; Jossberger et al., 2020). Digital tools in the classroom can also enhance students’ motivation and involvement (Tække & Paulsen, 2021). Smartphones, for example, can provide motivational support in learning contexts (Pachler et

123

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

al., 2010). At the same time, negative factors such as distraction, boredom, or costs can hinder these positive effects. Moreover, online learning platforms provide students with tools to set goals, track improvements, and reflect on their learning strategies, that enables them to develop metacognitive awareness and capabilities, and hence, their methodical competence (Dabbagh & Kitsantas, 2012; Ravindran et al., 2022). Work diaries and online learning modules help students to reflect on their ways of doing things, set personal goals, and monitor their learning, thereby helping to better understand their professional, personal, and methodical competencies (Quesada-Pallarès et al., 2019; Rausch, 2011). Private Use of Digital Media Digital media can also help develop students’ social and emotional competencies outside of the formal educational environment (McNaughton & Jesson, 2023). Social media platforms and other forms of communication technologies (Lahuerta-Otero et al., 2019; Tzur et al., 2022) can help students to become more self-motivated through interaction and social exchange. Connectivity among digital communities can promote engagement in learning. In addition, collaborative digital activities outside the classroom can help students develop communication and teamwork skills (McNaughton & Jesson, 2023). Working together through digital collaboration strategies, such as group projects or online exchanges, may strengthen students’ ability to coordinate tasks, share knowledge, and cooperate with others. Digital platforms can also support peer learning and mutual support, fostering the development of interpersonal skills in vocational and technical professions (McNaughton & Jesson, 2023). Integrated Impact Integrating digital media in VET can support the development of students’ SLCs in meaningful ways and help prepare them for both vocational practice and ongoing learning (Pylväs et al., 2022). When digital tools are used in both formal and informal learning contexts, they contribute to a learning environment that addresses multiple dimensions of SLC simultaneously. Professional competence can be strengthened through access to current, domain-specific digital resources and simulations. Methodical competence is supported by tools that facilitate planning, monitoring, and reflection on learning processes. Personal competence is encouraged by the greater autonomy required by digital learning environments (Barak, 2010; Dabbagh & Kitsantas, 2012). Digital media also plays a crucial role in social and emotional skills development. Online platforms and social media offer opportunities for interaction, communication and cooperation with others and can help strengthen our interpersonal skills and a sense of belonging (McNaughton & Jesson, 2023; Tzur et al., 2022). These effects are not limited to formal learning environments but also emerge in learners’ everyday use of digital media, showing that SLC development is influenced in different contexts.

124

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

As digital skills become more relevant in the labor market, it is critical that VET students develop technical skills as well as the skills to manage their learning in the process (Eickelmann et al., 2019; UNESCO, 2024). This is why digital media in VET can not only contribute to immediate learning outcomes but also to long-term professional development and continued learning (Cattaneo et al., 2021; Pylväs et al., 2022). The Present Study As described above, incorporating digital media into teaching can affect distinct but related dimensions of SLC. First, digital media can contribute to the development of professional competence by providing access to diverse, up-to-date information and enabling more practiceoriented, application-based learning experiences. In particular, in vocational contexts, digital tools allow learners to engage with authentic tasks, simulations, and problem-based scenarios, which support the acquisition and application of domain-specific knowledge (Cattaneo et al., 2021; Gräsel et al., 2020). At the same time, these environments require learners to actively connect new information with their prior knowledge, which is a central aspect of professional competence (Arnold et al., 2003). Second, digital learning environments are associated with methodical competence because they often require students to organize and regulate their own learning processes. E-learning platforms, online resources and digital assignments require students to plan their learning activities, choose appropriate strategies and monitor their progress. In that sense, digital media may contribute to metacognitive processes of planning, control and reflection (Barak, 2010; Dabbagh & Kitsantas, 2012; Jossberger et al., 2020). Digital work diaries and learning management systems can enhance these processes by providing structure and feedback (Quesada-Pallarès et al., 2019; Rausch, 2011). Third, digital media may also affect personal competence in motivation, autonomy, and persistence. Digital learning environments may also provide more flexible, personalized learning pathways that foster learners’ sense of autonomy and responsibility for their learning. This is closely related to the core aspects of personal competence such as self-efficacy and intrinsic motivation (Arnold et al., 2003; Deci & Ryan, 1993). Digital tools in learning contexts have been associated with increased engagement and participation, which may also support motivational processes (Pachler et al., 2010; Tække & Paulsen, 2021). But how such tools are employed in education should be clearly understood, as poorly designed digital learning environments may reduce motivation or result in disengagement. Taken together, the existing literature suggests that the use of digital media for school-related purposes is not limited to supporting a single dimension of SLC. Rather, it has the potential to simultaneously influence professional, methodical, and personal competences by combining access to knowledge, opportunities for self-regulation, and motivational support within learning processes. Therefore, we hypothesize:

125

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

(H1): There is a positive relationship between vocational students’ use of digital media for school-related purposes and their SLC in their professional, methodical, and personal competences. Further, the use of digital media for private purposes is a different, less structured context which may still contribute to the development of individual dimensions of competence for selflearning. Private digital media use is more self-initiated and is associated with everyday communication and social interaction, unlike school-related media use, which tends to be task and instruction-oriented. This is particularly true for the use of social media platforms where exchange, participation, and communication are key. Participation in these digital environments may allow the development of social competence through learners’ frequent engagement in interacting with others, coordinating activities, and communicating in these spaces. Such interactions can promote skills such as communication, cooperation, and relationship management, which are key components of social competence (McNaughton & Jesson, 2023). Hence, collaborative digital activities in informal settings may also contribute to the development of teamwork and coordination skills that are relevant for vocational contexts. Private digital media use may also influence emotional competence. Digital communication often requires individuals to recognize, interpret, and regulate emotions in interactions with others, for example, in online discussions or social media exchanges. Such experiences can support the development of emotional awareness, empathy, and emotion regulation. These are important aspects of emotional competence (Arnold et al., 2003). In addition, digital interaction and social exchange can strengthen feelings of belonging and connectedness, which have been associated with higher levels of motivation and engagement (Lahuerta-Otero et al., 2019; Tzur et al., 2022). But these effects are not necessarily all positive, and private use of digital media can also include distractions or negative experiences. Current research however suggests that the interactive and socially embedded nature of private digital media use may offer opportunities for the development of social and emotional competencies. Hence, we hypothesize: (H2): Use of digital media for private purposes is positively correlated with vocational students’ emotional and social competencies. These hypotheses aim to examine the relationship between vocational students’ digital media use and their SLCs. Investigating these relationships may help improve our understanding of how digital media can support competence development in vocational education.

126

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Research Methods Sample and Procedure This study uses a quantitative cross-sectional design and collects data from vocational education students at one point in time. This is appropriate to assess the current state of the variables under study (Mohajan, 2020). The study was conducted in accordance with the ethical principles of the Declaration of Helsinki. Prior to data collection, we obtained official approval from the responsible supervisory authority of the participating vocational school (Schullandesamt Halle: State School Authority of Halle, Germany). All participants provided written informed consent before participating in the survey. Participation was voluntary, and all data were collected and processed anonymously. Students from a German vocational school were invited to participate, regardless of their training area or year of study. A total of N = 248 students from three training areas participated in the study. The sample included students from automotive technology, metal technology, and business and administration, with the largest group coming from business and administration (58.0%), followed by metal technology (31.5%) and automotive technology (8.5%). Participants were almost evenly distributed across the two age groups, with 49.6% aged 16–18 and 50.4% aged 18 or older. The gender distribution showed a higher proportion of male (57.7%) than female students (41.1%), while one participant identified as diverse and two participants did not provide this information. Most students were in their first year of training (66.1%), followed by the third year (25.0%) and the second year (8.1%). Overall, the sample covers different vocational fields and training stages, with a clear concentration in business and administration and first-year students (see Table 1).

127

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Table 1 Demographic Characteristics of the Sample Frequency

Percentage

248

100%

Male Female Diverse Information not available

143 102 1 2

57.7% 41.1% 0.4% 0.8%

< 16 16–18 > 18

0 123 125

0% 49.6% 50.4%

Automotive Technology Metal Technology Business and Administration Information not available

21 78 144 5

8.5% 31.5% 58.0% 2.0%

1st year 2nd year 3rd year information not available

164 20 62 2

66.1% 8.1% 25.0% 0.8%

N Gender

Age

Training Area

Training Year

Instruments The underlying study employs a comprehensive survey to assess SLCs, using an adapted version of Arnold’s SLC questionnaire (Arnold, 2003) and items based on the ICILS study framework (Eickelmann et al., 2019). The SLC section comprised several subscales addressing professional, methodical, personal, emotional, and social competences. The items were adapted to the vocational school context. The final questionnaire also included questions on sociodemographic information, including participants’ gender, age, professional training field, and current year of training. Overall, the survey focused on evaluating students’ SLCs and examining how vocational students use digital media for school-related and private purposes. During this assessment, participants were asked to rate their agreement with statements related to learning behavior, motivation, emotions, and media use. The self-evaluation employed a five-point Likert scale ranging from “Strongly Disagree” to “Strongly Agree”. This approach enabled a differentiated assessment of students’ perceptions of their competences and learning behaviors in the context of vocational education.

128

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

A reliability analysis (Table 2) was used to check internal consistency of competency scales. Some items were excluded to improve reliability, mainly because the questionnaire was originally not developed for vocational students. The EC dimension was retained despite the low reliability (α = 0.47), as the scale emerges from a formative measurement structure of the model; this concept structure where each dimension independently contributes to the concept is the case where internal consistency is not so important (Gruijters et al., 2021). Emotional Competence includes self-perception and external perception, which likely explains the low reliability, but both are crucial for fully understanding this competence. Nevertheless, we kept EC dimension due to its necessity to complete the model and its overall importance. The reliability for the others was adequate, confirming the instrument’s internal consistency for this target group (Döring, 2023; Streiner, 2003). Table 2 Cronbach’s Alpha Values for Self-Learning Competence (SLC) Dimension

Example Item

Items

Professional Competence (PRC)

"I already knew something about my training area before I started my apprenticeship."

.73

Methodical Competence (MC)

"Before learning, I think about the order in which I work through the topics."

.68

Personal Competence (PEC)

"I want to finish the work I have started."

.77

Emotional Competence (EC)

"I feel accepted by my classmates."

.47

Social Competence (SC)

"I like discussing the subject matter with others."

.65

Note. α ≥ 0.9 = excellent; 0.7 ≤ α < 0.9 = good; 0.6 ≤ α < 0.7 = acceptable; 0.5 ≤ α < 0.6 = poor; α ≤ 0.5 = unacceptable (Streiner, 2003).

The operationalization of students’ digital media use in this study was formed by integrating concepts and measurement instruments from the ICILS study (Eickelmann et al., 2019). Several items focusing on the use of digital media were adapted and integrated into the research context of this study. This methodological decision enabled a well-founded, empirically supported cross-sectional assessment of participants’ use of digital media for school- and private purposes. Since the measurement of vocational school students’ digital media use for school-related and private purposes involved variables across different subscales, several tests were carried out to establish whether these items measure the underlying concepts consistently and reliably. First an exploratory factor analysis was conducted with which all the 12 items were grouped in four clusters, one labeled as "Use for lesson preparation"(three items), another “Use for private purposes” (two items), “Use of software programs” (three items) and finally “Use for other

129

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

school-related purposes” (four items). The two private-use items captured media use for nonschool-related purposes (social contact and self-presentation online), distinguishing them in content from the school-related items. Following this, a reliability analysis was performed to further refine the scales and optimize internal consistency. According to Table 3, the resulting Cronbach’s α values ranged from .72 to .77, which indicates an acceptable to good internal consistency across all four subscales (Streiner, 2003). This four-factor structure maps onto the two theoretically derived categories underlying the hypotheses: three of the factors represent subdimensions of school-related use, while use for private purposes constitutes the private use of digital media. Accordingly, H1 is examined through the paths from the three school-related subdimensions to professional, methodical, and personal competence, and H2 through the paths from use for private purposes to emotional and social competence. H1 is considered supported to the extent that the school-related subdimensions show the hypothesized positive relations. All remaining paths in the structural model were examined on an exploratory basis. Table 3 Cronbach’s Alpha Values for Digital Media Use Dimension

Items Included

Items

Lesson Preparation

Preparing essays and reports; Preparing presentations; Communicating with classmates

.73

Software Programs

Word processing programs, Presentation programs, Spreadsheet programs

.77

Private Purposes

Meeting new people online, Showing identity on social networks

.74

Other School-related Purposes

Searching for information for homework, completing school tasks, using digital media at school/outside for school purposes

.72

Note. α ≥ 0.9 = excellent; 0.7 ≤ α < 0.9 = good; 0.6 ≤ α < 0.7 = acceptable; 0.5 ≤ α < 0.6 = poor; α ≤ 0.5 = unacceptable (Streiner, 2003).

130

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Analysis We employed a comprehensive set of statistical analyses to investigate the research hypotheses guiding this study. Initially, correlation analyses were conducted in SPSS to examine associations between digital media use and dimensions of SLCs. Taking these into account a structural equation model in the form of a path-analysis with observed variables was set up with Jamovi (Navarro & Foxcroft, 2025). This approach allowed us to test all the relationships between various types of digital media use and SLC dimensions at the same time, and to allow unearthing possibly predictive relationships among them. Results Descriptive Analysis Descriptive statistics were calculated for SLC and digital media use among vocational students. The following tables summarize the distributions, mean values, and variability across the different competence dimensions and digital media use contexts. Table 4 Descriptive Statistics for Self-Learning Competence (SLC) SLC

Mdn

Min

Max

PRC

237

3.01

3.00

0.95

1.00

5.00

235

3.04

3.00

0.75

1.00

5.00

PEC

232

3.71

3.67

0.66

1.50

5.00

229

3.52

3.50

0.65

1.50

5.00

228

3.26

3.33

0.85

1.00

5.00

Note. PRC = professional competence; MC = methodical competence; PEC = personal competence; EC = emotional competence; SC = social competence; n = number of participants; M = mean; Mdn = median; SD = standard deviation; Min = Minimum; Max = Maximum

The descriptive statistics show that vocational students reported moderate to high levels across all competence dimensions (Table 4), with personal competence showing the highest mean score (M = 3.71).

131

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Table 5 Descriptive Statistics for Digital Media Use (DM) DM Subscale

Mdn

Min

Max

Lesson Preparation

227

3.83

4.00

.91

1.00

5.00

Software Programs

225

2.53

2.33

1.16

1.00

5.00

Private Purposes

225

2.50

1.14

1.50

5.00

Other School-related

227

3.32

3.25

.79

1.50

5.00

Note. n = number of participants; M = mean; Mdn = median; SD = standard deviation; Min = Minimum; Max = Maximum

The descriptive statistics for digital media use (DM) show that vocational students use digital media at varying levels across different contexts (see Table 5). Lesson preparation showed the highest mean value (M = 3.83), followed by other school-related purposes (M = 3.32), whereas software programs and private purposes were used less frequently. Correlation Analysis The hypotheses are analyzed in a two-step procedure. First, a correlation analysis was conducted between the components of the SLC and the aggregated variables for digital media use (see Table 6). Table 6 Correlation Analysis: SLC and Digital Media Use Scale

Lesson Preparation

Private Purposes

Software Programs

Other Schoolrelated

PRC

.151*

-.014

.103

.110

.134**

-.029

-.067

.187**

PEC

.204**

.105

-.039

.299**

.335**

.152*

.098

.335**

.432**

.100

.033

.338**

Note. PRC = professional competence; MC = methodical competence; PEC = personal competence; EC = emotional competence; SC = social competence. *p < .05. **p < .01.

The correlation analysis shows relationships among various SLCs and different forms of digital media use. The results indicate that professional, methodical, and personal competencies are linked to digital media use in the classroom. A small but significant positive correlation exists between professional competence and digital media use for lesson preparation (r = .151, p < .05). Methodical competence correlates with both lesson preparation (r = .134, p < .01) and school-related purposes (r = .187, p < .01). In contrast, personal competence shows a moderate

132

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

correlation with lesson preparation (r = .204, p < .01) and school-related purposes (r = .299, p < .01). There is a small positive correlation between emotional competence and digital media use for private purposes (r = .152, p < .05) and a moderate correlation with school-related purposes (r = .335, p < .01). Social competence is positively and significantly correlated with lesson preparation (r = .432, p < .01) and other school-related digital media use (r = .338, p < .01). These bivariate results provide preliminary insights into hypothesized relationships. Several school-related subdimensions correlate positively with professional, methodical, and personal competence, while private use only correlates positively with emotional competence. However, as correlations do not account for the interrelations among the media-use dimensions, the structural equation model provides the primary test of the hypotheses. Structural Equation Modeling (SEM) A structural equation model in the form of a path analysis was conducted to further examine the relationships between different forms of digital media use and the dimensions of SLC. The model enabled the analysis of multiple associations between variables within a single analytical framework. Because the specified path model was just-identified (df = 0), traditional fit indices such as CFI, RMSEA, and SRMR were not informative and were therefore not interpreted (Kline, 2023). The results of the path analysis are presented in Table 7 and Figure 2.

133

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Table 7 Structural Equation Modeling (SEM) Parameter Estimates Parameter

PRC ← Lesson Preparation

.126

.075

.12

1.675

.094

PRC ← Software Programs

-.059

.057

-.072 -1.038

.299

PRC ← Private Purposes

.059

.055

.07

1.068

.286

PRC ← Other School-related

.086

.092

.071

.939

.348

MC ← Lesson Preparation

.106

.056

.129

1.879

.060

MC ← Software Programs

-.055

.042

-.085 -1.291

.197

MC ← Private Purposes

-.098

.041

-.148 -2.366

.018

MC ← Other School-related

.276

.069

.292

4.020 < .001

PEC ← Lesson Preparation

.077

.050

.106

1.532

.126

PEC ← Software Programs

-.010

.038

-.017

-.260

.795

PEC ← Private Purposes

-.066

.037

-.113 -1.800

.073

PEC ← Other School-related

.233

.061

.279

3.813 < .001

EC ← Lesson Preparation

.165

.048

.232

3.453 < .001

EC ← Software Programs

.009

.036

.017

.260

.795

EC ← Private Purposes

.003

.035

.006

.097

.923

EC ← Other School-related

.185

.058

.226

3.182

.001

SC ← Lesson Preparation

.340

.060

.366

5.666 < .001

SC ← Software Programs

-.038

.045

-.052

-.843

.399

SC ← Private Purposes

-.059

.044

-.079 -1.344

.179

SC ← Other School-related

.227

.073

.212

.002

Professional Competence (PRC)

Methodical Competence (MC)

Personal Competence (PEC)

Emotional Competence (EC)

Social Competence (SC)

3.109

Note. B = estimate; SE = standard error; β = standardized coefficient; z = z-statistic; p = significance value.

134

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Figure 2 Path Diagram of the Structural Equation Model Showing Standardized Coefficients of Significant Paths

Note. Only significant paths (p < .05) are displayed; standardized coefficients (β) are shown. Full parameter estimates are reported in Table 7.

Figure 2 depicts the significant paths of the structural model. Significant associations can be found primarily for lesson preparation and other school-related media use, whereas private media use showed only a single negative association with methodical competence. Professional competence and software program use are not displayed because no significant paths emerged for these variables. As described above, the SEM showed no significant relationships between Professional Competence (PRC) and any form of digital media use. Lesson preparation, software program use, private media use, and other school-related media use were not significantly associated with professional competence within the present model. For Methodical Competence (MC), the SEM showed a significant positive relationship with other school-related media use (β = .292, p < .001) and a significant negative relationship with private media use (β = -.148, p = .018). This suggests that higher levels of methodical competence are associated with more frequent school-related media use and less frequent private media use. No significant relationships were found for lesson preparation or software program use. For Personal Competence (PEC), only other school-related media use showed a significant positive relationship (β = .279, p < .001). The previously observed association with lesson preparation was no longer significant in the SEM. Emotional Competence (EC) was positively associated with both lesson preparation (β = .232, p < .001) and other school-related media use (β = .226, p = .001). In contrast, no significant relationships emerged for software program use or private media use.

135

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

For Social Competence (SC), the SEM identified significant positive relationships with lesson preparation (β = .366, p < .001) and with other school-related media use (β = .212, p = .002), consistent with the correlation analysis. Regarding the hypothesized paths, H1 received partial support, whereas H2 was not supported. Specifically, use of digital media for other school-related purposes was positively associated with methodical and personal competence, while no significant relationships were found for professional competence or the remaining school-related media use dimensions. Contrary to H2, use for private purposes was not positively associated with emotional or social competence. Notably, the small positive correlation between private use and emotional competence observed in the correlation analysis did not hold in the structural model, in which the mediause dimensions were considered simultaneously. Beyond the hypothesized relationships, the exploratory analyses identified several additional significant paths. Use for lesson preparation and use for other school-related purposes were positively associated with emotional and social competence, whereas use for private purposes showed a negative association with methodical competence. As these relationships were not hypothesized a priori, they should be interpreted with caution and require confirmation in future research. Overall, the SEM provides a more differentiated picture of the relationships between digital media use and SLC than the correlation analyses alone. Discussion Professional Competence and School-Related Digital Media Use Contrary to H1, professional competence was not associated with school-related digital media use. One possible explanation is that vocational education relies on hands-on practical experience, which cannot be fully replaced by digital media. This interpretation is consistent with previous research highlighting the limitations of digital tools for developing occupationspecific practical skills (Cattaneo et al., 2021). Methodical Competence and School-Related Digital Media Use Methodical competence showed the clearest association with school-related digital media use, supporting the assumption that digital learning environments might foster competences related to planning, organizing, and regulating learning processes. This could be due to school-related media-use activities assessed in this study inherently require learners to structure and monitor their learning. This interpretation aligns with the conceptualization of methodical competence as metacognitive control (Arnold et al., 2003; Jossberger et al., 2020). Also, digital learning environments may promote these processes by providing opportunities for planning, feedback, and reflection (Dabbagh & Kitsantas, 2012). The negative association with private media use further suggests that students with stronger methodical competence may regulate their media

136

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

use more strategically and avoid potentially distracting digital activities, similar to Barak’s (2010) findings on self-regulation and media use. Personal Competence and School-Related Digital Media Use The positive association between personal competence and school-related digital media use suggests that digital learning environments may benefit learners with high levels of motivation, perseverance, and self-efficacy. This interpretation is consistent with self-determination theory, which emphasizes autonomy and self-regulation as key drivers of motivated learning (Deci & Ryan, 1993). School-related digital media use often involves collaborative and self-directed activities that may strengthen learners’ sense of ownership over their learning. By contrast, no association can be found for the use of standard software programs, indicating that technical tools alone are unlikely to foster personal competence unless they are embedded in meaningful learning activities that support autonomy and engagement (Pachler et al., 2010). Emotional Competence and Digital Media Use Contrary to H2, emotional competence was not associated with private digital media use when the other media-use dimensions were controlled for. This finding suggests that everyday digital interaction alone may not be sufficient to support emotional competence. Instead, the exploratory associations with school-related media use indicate that emotionally competent learners may be more likely to engage with digital learning environments that require persistence, emotion regulation, and sustained engagement. This interpretation aligns with research showing that emotionally competent learners are generally better able to cope with academic challenges and maintain motivation during learning processes (Schiefele & Pekrun, 1996). Social Competence and Digital Media Use Private digital media use was also not associated with higher levels of social competence. A possible explanation is that many private online interactions remain relatively informal and may not provide sufficient opportunities to develop more complex interpersonal competences such as teamwork, conflict resolution, or goal-oriented collaboration. Instead, the positive associations with school-related digital media use suggest that these competences are more likely to develop in structured learning environments where collaboration serves a shared academic purpose. McNaughton and Jesson’s (2023) research also emphasizes that collaborative learning activities, rather than social media use itself, foster the development of social competence in vocational education. Recommendations Based on the findings of this study, several implications for educational practice and future research can be derived.

137

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Vocational schools should integrate digital media into structured learning activities that explicitly promote self-directed learning. Rather than focusing on digital tools themselves, schools should prioritize their pedagogical integration into goal-oriented learning tasks. This recommendation reflects previous research showing that the effectiveness of digital media depends primarily on their pedagogical design rather than on the technology itself (Bläsche, 2018; Higgins et al., 2012; Tamim et al., 2011). Second, vocational education should promote a more reflective and purposeful use of digital media by explicitly addressing media literacy and self-regulation within the curriculum. Supporting students in distinguishing between productive and distracting forms of media use may help them use digital technologies more effectively for learning. Previous research has consistently highlighted self-regulation as a key factor linking digital media use and successful learning outcomes (Dabbagh & Kitsantas, 2012; Zimmerman, 2002). Finally, teacher education and professional development should place greater emphasis on competence-oriented approaches to digital media integration. Beyond technical skills, teachers need pedagogical knowledge to design learning environments in which digital media support self-directed learning. This recommendation is consistent with the TPACK framework (Mishra & Koehler, 2006), which emphasizes the integration of technological, pedagogical, and content knowledge. Evidence suggests that TPACK-oriented professional development can improve technology-supported instruction (Tondeur et al., 2012; Voogt et al., 2013). Conclusion This study contributes to the growing body of research on digital media use in vocational education by demonstrating that the educational value of digital media depends less on their use than on how they are integrated into learning processes. By differentiating between distinct forms of digital media use and multiple dimensions of SLC, the study provides a detailed perspective on the role of digital media in competence-related outcomes. Moreover, the findings highlight the importance of considering not only whether digital media are used, but also the pedagogical purposes they serve within vocational education. The findings underline that vocational learning continues to depend on authentic practical experiences, which cannot be fully replaced by digital technologies. Instead, the educational value of digital media appears to lie in their potential to support learning-related competences, such as self-regulation, motivation, and learning strategies. Consequently, vocational education should focus less on providing access to digital tools and more on designing learning environments in which digital media are purposefully integrated to support SLCs. In this sense, digital media should be understood as complementing rather than replacing practical learning experiences. Several limitations should be considered when interpreting these findings. First, the crosssectional design does not allow conclusions about causal relationships and does not capture changes over time. Second, all variables were assessed by self-report, making social

138

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

desirability and inaccurate self-assessments possible. Third, the study did not distinguish clearly between formal and informal contexts of digital media use. Fourth, the relatively low reliability of the emotional competence scale may have affected the corresponding findings. Finally, several reported associations emerged from exploratory analyses and therefore require replication. Future research should address these limitations through longitudinal or mixedmethods designs, objective measures of digital media use and competence development, and more diverse educational contexts.

Acknowledgements The authors would like to thank the participating vocational school and its students for taking part in this study. Declaration of Generative AI and AI-assisted Technologies We also acknowledge the use of AI-supported tools, including ChatGPT, Grammarly, and Claude, during the writing process. These tools were used only to support language, clarity, and readability. All the content, analyses, interpretations, and conclusions are the responsibility of the authors.

139

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

References Alam, T. M., Stoica, G. A., Sharma, K., & Özgöbek, Ö. (2025). Digital technologies in the classrooms in the last decade (2014–2023): A bibliometric analysis. Frontiers in Education, 10, Article 1533588. https://doi.org/10.3389/feduc.2025.1533588 Arnold, R., & Gómez Tutor, C. (2006). Grundlinien einer Ermöglichungsdidaktik: Bildung ermöglichen – Vielfalt gestalten [Outlines of an enabling didactics: Enabling education – shaping diversity]. ZIEL. Arnold, R., Gómez Tutor, C., & Kammerer, J. (2003). Selbstlernkompetenzen als Voraussetzungen einer Ermöglichungsdidaktik – Anforderungen an Lehrende [Selflearning competences as prerequisites for enabling didactics – requirements for teachers]. In R. Arnold & I. Schüßler (Eds.), Ermöglichungsdidaktik in der Erwachsenenbildung [Enabling didactics in adult education] (pp. 108–119). Schneider. Backes-Gellner, U., & Lehnert, P. (2023). Berufliche Bildung als Innovationstreiber: Ein lange vernachlässigtes Forschungsfeld [Vocational education as a driver of innovation: A long neglected field of research]. Perspektiven der Wirtschaftspolitik, 24(1), 85–97. https://doi.org/10.1515/pwp-2022-0036 Barak, M. (2010). Motivating self-regulated learning in technology education. International Journal of Technology and Design Education, 20(4), 381–401. https://doi.org/10.1007/s10798-009-9092-x Barz, N., Benick, M., Dörrenbächer-Ulrich, L., & Perels, F. (2024). Students’ acceptance of e-learning: Extending the technology acceptance model with self-regulated learning and affinity for technology. Discover Education, 3, Article 114. https://doi.org/10.1007/s44217-024-00195-7 Beyer, L. (2020). Selbstlernkompetenz in Präsenz und semi-virtuellen Lehrkonzepten [Selflearning competence in face-to-face and semi-virtual teaching concepts]. Empirische Evaluationsmethoden, 24, 55–70. ZeE Verlag. Bläsche, A. (2018). Arbeiten und Qualifizieren in digitalen Zeiten: Strategische Überlegungen für eine Berufliche Bildung 4.0 [Working and qualifying in digital times: Strategic considerations for vocational education 4.0]. In F. Schröder (Ed.), Auf dem Weg zur digitalen Aus- und Weiterbildung von morgen: Ergebnisse des Berliner Modells „Zusatzqualifikationen für digitale Kompetenzen“ [On the way to tomorrow's digital initial and continuing training: Results of the Berlin model "Additional qualifications for digital skills"] (pp. 1–128). wbv Publikation. https://doi.org/10.3278/6004656w Bremer, C., & Antony, I. (2013). Einsatz digitaler Medien für den lernerzentrierten Unterricht: Konzeption und Evaluation der Lehrerfortbildung „Lernkompetenz entwickeln, individuell fördern“ [Use of digital media for learner-centered teaching: Conception and evaluation of the teacher training "Developing learning competence, promoting individually"]. In Mediendidaktik in der Weiterbildung [Media didactics in continuing education] (Vol. 72). https://www.pedocs.de/volltexte/2018/16148/pdf/MidW_72_Bremer_Antony_Einsatz _digitaler_Medien.pdf

140

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Cattaneo, A. A., Antonietti, C., & Rauseo, M. (2021). How digitalised are vocational teachers? Assessing digital competence in vocational education and looking at its underlying factors. Computers & Education, 176, 104358. https://doi.org/10.1016/j.compedu.2021.104358 Dabbagh, N., & Kitsantas, A. (2012). Personal learning environments, social media, and selfregulated learning: A natural formula for connecting formal and informal learning. The Internet and Higher Education, 15(1), 3–8. https://doi.org/10.1016/j.iheduc.2011.06.002 Deci, E. L., & Ryan, R. M. (1993). Die Selbstbestimmungstheorie der Motivation und ihre Bedeutung für die Pädagogik [Self-determination theory of motivation and its significance for pedagogy]. Zeitschrift für Pädagogik, 39(2), 223–238. https://doi.org/10.25656/01:11173 Döring, N. (2023). Forschungsmethoden und Evaluation in den Sozial- und Humanwissenschaften [Research methods and evaluation in the social and human sciences] (6th ed.). Springer. https://doi.org/10.1007/978-3-662-64762-2 Eickelmann, B. (2017). Herausforderungen und Perspektiven: Lernende Schule in der digitalen Welt [Challenges and perspectives: Learning school in the digital world]. Lernende Schule, 20(79), 4–9. Eickelmann, B. (2019). Schule und Lernen unter Bedingungen der Digitalisierung: Wie können Potenziale digitaler Medien für die Entwicklung der Lernkultur in Sekundarschulen genutzt werden? [Schooling and learning under conditions of digitalization: How can the potential of digital media be used to develop learning culture in secondary schools?]. Pädagogik, 71(3), 34–37. Eickelmann, B., Bos, W., Gerick, J., Goldhammer, F., Schaumburg, H., Schwippert, K., Senkbeil, M., & Vahrenhold, J. (Eds.). (2019). ICILS 2018 #Deutschland – Computer- und informationsbezogene Kompetenzen von Schülerinnen und Schülern im zweiten internationalen Vergleich und Kompetenzen im Bereich Computational Thinking [ICILS 2018 #Germany – Computer and information literacy of students in a second international comparison and competencies in computational thinking]. Waxmann. Eickelmann, B., & Gerick, J. (2017). Lehren und Lernen mit digitalen Medien [Teaching and learning with digital media]. Schulmanagement Handbuch, 164(4), 54–81. European Centre for the Development of Vocational Training (CEDEFOP). (2022). The future of vocational education and training in Europe. Volume 3. The influence of assessments on vocational learning. Publications Office. https://doi.org/10.2801/067378 Findeisen, S., & Wild, S. (2022). General digital competences of beginning trainees in commercial vocational education and training. Empirical Research in Vocational Education and Training, 14(1), 2. https://doi.org/10.1186/s40461-022-00130-w Fraillon, J., Ainley, J., Schulz, W., Friedman, T., & Gebhardt, E. (2014). Preparing for life in a digital age: The IEA International Computer and Information Literacy Study international report. Springer. https://doi.org/10.1007/978-3-319-14222-7 Gräsel, C., Schledjewski, J., & Hartmann, U. (2020). Implementation digitaler Medien als Schulentwicklungsaufgabe [Implementation of digital media as a school development task]. Zeitschrift für Pädagogik, 66(2), 208–224. https://doi.org/10.3262/ZP2002208

141

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Gruijters, S. L. K., Fleuren, B. P. I., & Peters, G.-J. Y. (2021). Crossing the seven Cs of internal consistency: Assessing the reliability of formative instruments [Preprint]. PsyArXiv. https://doi.org/10.31234/osf.io/qar39 Haasler, S. R. (2020). The German system of vocational education and training: Challenges of gender, academisation and the integration of low-achieving youth. Transfer: European Review of Labour and Research, 26(1), 57–71. Hannack, E., & Wucherpfennig, D. (2018). Gute Arbeit gestalten für die digitale Zukunft [Shaping good work for the digital future]. In F. Schröder (Ed.), Auf dem Weg zur digitalen Aus- und Weiterbildung von morgen: Ergebnisse des Berliner Modells „Zusatzqualifikationen für digitale Kompetenzen“ [On the way to tomorrow's digital initial and continuing training: Results of the Berlin model "Additional qualifications for digital skills"] (pp. 103–107). wbv Publikation. https://doi.org/10.3278/6004656w Higgins, S., Xiao, Z., & Katsipataki, M. (2012). The impact of digital technology on learning: A summary for the Education Endowment Foundation. Durham University. Hemmler, Y. M., & Ifenthaler, D. (2024). Self-regulated learning strategies in continuing education: A systematic review and meta-analysis. Educational Research Review, 45, Article 100629. https://doi.org/10.1016/j.edurev.2024.100629 Hugo, J., Fehrmann, R., & Ud-Din, S. (Eds.). (2022). Digitalisierungen in Schule und Bildung als gesamtgesellschaftliche Herausforderung: Perspektiven zwischen Wissenschaft, Praxis und Recht [Digitalization in schools and education as a challenge for society as a whole: Perspectives between science, practice, and law]. Waxmann. Jossberger, H., Brand-Gruwel, S., van de Wiel, M. W. J., & Boshuizen, H. P. A. (2020). Exploring students’ self-regulated learning in vocational education and training. Vocations and Learning, 13(1), 131–158. https://doi.org/10.1007/s12186-019-09232-1 Kline, R. B. (2023). Principles and practice of structural equation modeling (5th ed.). Guilford Press. https://doi.org/10.1037/14136-000 Knowles, M. S. (1975). Self-directed learning: A guide for learners and teachers. Association Press. Kovalchuk, V. I., & Sheludko, I. V. (2019). Implementation of digital technologies in training the vocational education pedagogues as a modern strategy for modernization of professional education. Annales Universitatis Paedagogicae Cracoviensis. Studia ad Didacticam Biologiae Pertinentia, 9, 122–138. https://doi.org/10.24917/20837276.9.13 Lahuerta-Otero, E., Cordero-Gutiérrez, R., & Izquierdo-Álvarez, V. (2019). Using social media to enhance learning and motivate students in the higher education classroom. In L. Uden, D. Liberona, G. Sanchez, & S. Rodríguez-González (Eds.), Communications in Computer and Information Science: Learning Technology for Education Challenges (Vol. 1011, pp. 351–361). Springer. https://doi.org/10.1007/978-3-030-20798-4_30 McNaughton, S., & Jesson, R. (2023). How a digital intervention in schools contributed to students’ social and emotional skills, and impacted writing. New Zealand Journal of Educational Studies, 58(2), 361–377. https://doi.org/10.1007/s40841-023-00296-1

142

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Mejeh, M., & Held, T. (2022). Understanding the development of self-regulated learning: An intervention study to promote self-regulated learning in vocational schools. Vocations and Learning, 15, 531–568. https://doi.org/10.1007/s12186-022-09298-4 Mishra, P., & Koehler, M. J. (2006). Technological pedagogical content knowledge: A framework for teacher knowledge. Teachers College Record, 108(6), 1017–1054. https://doi.org/10.1111/j.1467-9620.2006.00684.x Mohajan, H. (2020). Quantitative research: A successful investigation in natural and social sciences. Journal of Economic Development, Environment and People, 9(4). https://doi.org/10.26458/jedep.v9i4.679 Müller, H., & Nannen-Gethmann, F. (2020). Berufliche Qualifizierung 4.0 – Konzepte und Ziele für die gewerblichen Berufe [Vocational qualification 4.0 – Concepts and goals for industrial professions]. In T. Vollmer, T. Karges, T. Richter, B. Schlömer, & S. Schütt-Sayed (Eds.), Digitalisierung mit Arbeit und Berufsbildung nachhaltig gestalten [Shaping digitalization sustainably through work and vocational education] (pp. 73–84). Navarro, D., & Foxcroft, D. (2025). Learning statistics with jamovi. Open Book Publishers. https://doi.org/10.11647/obp.0333 Pachler, N., Bachmair, B., & Cook, J. (2010). Mobile learning: Structures, agency, practices. Springer. Panadero, E. (2017). A review of self-regulated learning: Six models and four directions for research. Frontiers in Psychology, 8, 422. https://doi.org/10.3389/fpsyg.2017.00422 Pilz, M., & Fürstenau, B. (2019). Duality and learning fields in vocational education and training: Pedagogy, curriculum, and assessment. In D. Guile & L. Unwin (Eds.), The Wiley handbook of vocational education and training (pp. 311–327). Wiley. https://doi.org/10.1002/9781119098713.ch16 Pylväs, L., Li, J., & Nokelainen, P. (2022). Professional growth and workplace learning. In C. Harteis, D. Gijbels, & E. Kyndt (Eds.), Research approaches on workplace learning (pp. 137–155). Springer. https://doi.org/10.1007/978-3-030-89582-2_6 Quesada-Pallarès, C., Sánchez-Martí, A., Ciraso-Calí, A., & Pineda-Herrero, P. (2019). Online vs. classroom learning: Examining motivational and self-regulated learning strategies among vocational education and training students. Frontiers in Psychology, 10, 2795. https://doi.org/10.3389/fpsyg.2019.02795 Ramin, P. (2022). Digital competence and future skills: How companies prepare themselves for the digital future. Hanser. https://doi.org/10.3139/9783446474284 Rausch, A. (2011). Erleben und Lernen am Arbeitsplatz in der betrieblichen Ausbildung [Experience and learning in the workplace during company training]. Springer. https://doi.org/10.1007/978-3-531-93199-9 Ravindran, L., Ridzuan, I., & Wong, B. E. (2022). The impact of social media on the teaching and learning of EFL speaking skills during the COVID-19 pandemic. In International Academic Symposium of Social Science 2022 (p. 38). MDPI. https://doi.org/10.3390/proceedings2022082038

143

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Schiefele, U., & Pekrun, R. (1996). Psychologische Modelle des fremdgesteuerten und selbstgesteuerten Lernens [Psychological models of externally directed and selfdirected learning]. In F. E. Weinert (Ed.), Psychologie des Lernens und der Instruktion [Psychology of learning and instruction] (Enzyklopädie der Psychologie, Vol. 2, pp. 249–278). Hogrefe. Schulte, S., Richter, T., & Grantz, T. (2014). Digital media as support for technical vocational training: Expectations and research results of the use of Web2.0. International Journal of Advanced Corporate Learning (iJAC), 7(3), 29. https://doi.org/10.3991/ijac.v7i3.4013 Seever, F., & Schacher, P. (2022). Potenziale zum Erwerb von digitalisierungsbezogenen Kompetenzen durch den Einsatz digitaler Medien im Geschichtsunterricht [Potential for acquiring digitalization-related skills through the use of digital media in history teaching]. In J. Hugo, R. Fehrmann, & S. Ud-Din (Eds.), Digitalisierungen in Schule und Bildung als gesamtgesellschaftliche Herausforderung: Perspektiven zwischen Wissenschaft, Praxis und Recht [Digitalization in schools and education as a challenge for society as a whole: Perspectives between science, practice, and law] (pp. 155 ff.). Waxmann. Streiner, D. L. (2003). Starting at the beginning: An introduction to coefficient alpha and internal consistency. Journal of Personality Assessment, 80(1), 99–103. https://doi.org/10.1207/S15327752JPA8001_18 Tække, J., & Paulsen, M. (2021). A new perspective on education in the digital age: Teaching, media and Bildung. Bloomsbury Academic. https://doi.org/10.5040/9781350175426 Tamim, R. M., Bernard, R. M., Borokhovski, E., Abrami, P. C., & Schmid, R. F. (2011). What forty years of research says about the impact of technology on learning. Review of Educational Research, 81(1), 4–28. https://doi.org/10.3102/0034654310393361 Tondeur, J., van Braak, J., Sang, G., Voogt, J., Fisser, P., & Ottenbreit-Leftwich, A. (2012). Preparing pre-service teachers to integrate technology in education: A synthesis of qualitative evidence. Computers & Education, 59(1), 134–144. https://doi.org/10.1016/j.compedu.2011.10.009 Tzur, S., Katz, A., & Davidovitch, N. (2022). The impact of social networks on student motivation and achievement. In Globalisation, comparative education and policy research (pp. 119–140). https://doi.org/10.1007/978-3-030-92608-3_7 UNESCO. (2024). What you need to know about digital learning and transformation of education. https://www.unesco.org/en/digital-education/need-know Voogt, J., Fisser, P., Pareja Roblin, N., Tondeur, J., & van Braak, J. (2013). Technological pedagogical content knowledge – a review of the literature. Journal of Computer Assisted Learning, 29(2), 109–121. https://doi.org/10.1111/j.1365-2729.2012.00487.x Weinert, F. E. (2000). Lehren und Lernen für die Zukunft: Ansprüche an das Lernen in der Schule [Teaching and learning for the future: Demands on school learning]. Pädagogisches Zentrum Rheinland-Pfalz. Zimmerman, B. J. (2000). Attaining self-regulation: A social cognitive perspective. In M. Boekaerts, P. R. Pintrich, & M. Zeidner (Eds.), Handbook of self-regulation (pp. 13– 39). Academic Press. https://doi.org/10.1016/B978-012109890-2/50031-7

144

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Zimmerman, B. J. (2002). Becoming a self-regulated learner: An overview. Theory Into Practice, 41(2), 64–70. https://doi.org/10.1207/s15430421tip4102_2 Zumbach, J., & Astleitner, H. (2016). Effektives Lehren an der Hochschule: Ein Handbuch zur Hochschuldidaktik [Effective teaching at universities: A handbook for university didactics]. Kohlhammer. http://nbn-resolving.org/urn:nbn:de:bsz:24-epflicht-1291964

Corresponding author: Maxi Brausch-Böger Email: maxi.brausch@tum.de

145

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

146

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

AI Literacy and Convergence Competencies Among Pre-Service Teachers in Non-STEM Disciplines Juyoung Lee Seoul National University of Education, Republic of Korea

147

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Abstract Artificial intelligence (AI) is transforming education, requiring teachers to develop competencies that support AI convergence education across subject areas. Although AI literacy is increasingly emphasized in teacher education, empirical research on AI convergence education competencies among pre-service teachers, particularly in non-STEM disciplines, remains limited. This study examined levels of AI literacy and AI convergence education competencies among pre-service teachers and analyzed differences by academic major, gender, level of AI understanding, and prior AI-related educational experience. Survey data were collected from 112 pre-service teachers in humanities, social sciences, arts, and physical education at a teacher education institution in South Korea. Two validated instruments were used to measure AI literacy and AI convergence education competencies. Data were analyzed using descriptive statistics, one-sample t-tests (test value = 3.00), independent samples t-tests, and one-way ANOVA, with the false discovery rate controlled using the Benjamini-Hochberg procedure. Most subdomains of AI literacy and AI convergence education competencies scored significantly above the scale midpoint. Data literacy and basic AI knowledge did not differ significantly from it, and programming-related competencies were significantly lower. Significant differences emerged across majors, gender, and levels of AI understanding, indicating uneven preparedness. In contrast, within AI convergence education competencies, ethics and openness did not differ significantly by prior AI-related educational experience. These findings suggest the need for systematic, discipline-sensitive teacher education curricula that strengthen both foundational AI literacy and subject-specific convergence competencies in teacher preparation. Keywords: AI convergence education, AI literacy, non-STEM disciplines, pre-service teacher education, teacher competencies

148

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Artificial Intelligence (AI) has rapidly emerged as a core technology transforming multiple sectors of society, with particularly profound implications for educational paradigms (Maslej et al., 2025). According to UNESCO (2019), AI is expected to fundamentally reshape pedagogical methodologies, educational tools, learning environments, and teacher preparation programs. In response, many countries have begun embedding AI into national curricula and teacher education systems to prepare future generations for an AI-driven society. AI literacy, commonly defined as the ability to understand, interpret, critically assess, and ethically use AI technologies (Long & Magerko, 2020; OECD, 2019), has been conceptualized as a multidimensional construct encompassing technical, pedagogical, critical, and ethical dimensions in educational contexts (Ng et al., 2021). In South Korea, the Ministry of Education has positioned AI literacy as a core element of future competencies in its 2022 Revised National Curriculum. As part of the Digital-Based Educational Innovation Plan (Ministry of Education, 2023), the government has expanded AI-related coursework in teacher-training institutions. These initiatives reflect a growing consensus that pre-service teachers, regardless of subject specialization, should acquire the knowledge and skills necessary to integrate AI meaningfully into instructional practices. AI convergence education refers to the pedagogical integration of AI into subject-matter teaching, in which AI serves as a tool for learning within a discipline rather than as a standalone subject. For this integration to become embedded in classroom practice, teachers themselves must build robust competencies in AI convergence education. Such competencies involve not only technical knowledge but also the ability to design, implement, and evaluate lessons that incorporate AI in subject-relevant and ethically responsible ways. International studies suggest that teacher education programs increasingly recognize AI convergence education competencies as core professional capacities, although approaches to their development vary across contexts (Ng et al., 2021; Zawacki-Richter et al., 2019). Experiences within teacher education institutions are critical in shaping pre-service teachers’ dispositions and instructional approaches (Tondeur et al., 2012). Structured and sustained exposure to AI-focused pedagogy enhances teachers’ efficacy and fosters proactive attitudes toward AI convergence in education. Nonetheless, recent studies raise concerns regarding pre-service teachers' preparedness. According to Park (2021) and N. Kim and M. Kim (2024), pre-service teachers report only moderate levels of interest and understanding in AI, indicating a gap between national policy expectations and actual classroom readiness. Similar patterns have been reported internationally, with uneven levels of AI-related understanding and self-assessed expertise documented among pre-service teachers across diverse contexts (Dayagbil et al., 2025; Gamlem et al., 2026). This gap is especially pronounced among pre-service teachers, many of whom major in non-STEM subjects. Moon et al. (2021) and N. Kim and M. Kim (2024) suggest that today's pre-service teachers are part of a transitional generation who have not fully benefited from AI education policies, underscoring the need for focused educational interventions. Furthermore, a review of existing studies reveals a strong emphasis on elementary-level or information-education-focused pre-

149

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

service teachers, while empirical research examining AI literacy and AI convergence education competencies among pre-service teachers remains limited, particularly for those in humanities and social science disciplines (García-Peñalvo, 2023; Zhai et al., 2021). Accordingly, the purpose of this study is to examine AI literacy and AI convergence education competencies among pre-service teachers across diverse academic majors. The study aims to identify strengths and weaknesses in their preparation and to provide practical implications for teacher education programs. By analyzing how AI convergence education competencies vary by subject specialization, this research seeks to inform the development of responsive, futureoriented teacher education curricula that can better support pre-service teachers in becoming competent educators in an AI-integrated educational landscape. The findings are expected to provide empirical insights that can inform curriculum development and instructional design in teacher education for the AI era. Specifically, this study addresses the following research questions: •

• •

RQ1. What are the levels of AI literacy and AI convergence education competencies among pre-service teachers majoring in humanities, social sciences, arts, and physical education? RQ2. How do AI literacy levels differ according to academic major, gender, and level of AI understanding among pre-service teachers? RQ3. How do AI convergence education competencies differ according to prior AIrelated educational experience among pre-service teachers? Literature Review

AI Literacy of Pre-service Teachers Definitions of AI literacy agree on its breadth but differ on where its technical boundary lies. Long and Magerko (2020) characterize AI literacy as an extension of traditional literacy that encompasses understanding how AI works, interpreting data, using AI tools, and exercising ethical responsibility. This conception places the construct explicitly within reach of nonspecialists. OECD (2019) frames it in similar terms, as a civic capacity necessary for living freely in an age of coexistence with AI rather than as a technical specialization. Ng et al. (2021) retain a creation dimension in which learners build or modify AI artifacts, and Laru et al. (2025) report technical understanding and critical evaluation as separable contributors to the practical application of AI. The distinction between conceptions that limit AI literacy to evaluation and use and those that also include creation matters for teacher education. If AI literacy is understood as an evaluative and communicative capacity, programming ability sits at the periphery of the construct; if creation is included, programming becomes a criterion against which teachers can be judged underprepared. Measurement instruments derived from these frameworks inherit this

150

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

ambiguity, and studies that report low programming scores among teachers (e.g., Lim, 2023) are therefore difficult to interpret without knowing which conception the instrument assumes. National policy initiatives for AI in education have broadened expectations for AI literacy beyond information education specialists, requiring all teachers to hold foundational AI knowledge together with the pedagogical expertise to integrate AI into their own subject areas (Jeon et al., 2020; Tondeur et al., 2012). Empirical research on pre-service teachers’ AI literacy has grown quickly, but its coverage is uneven in ways that constrain what can be inferred. Work in the Korean context has concentrated on elementary pre-service and in-service teachers (Lee, 2020; Park, 2021; Ryu & Han, 2018), populations whose curricular obligations differ from those of secondary preservice teachers preparing to teach non-STEM subjects. Casal-Otero et al. (2023) document a comparable concentration in K-12 AI literacy research internationally, where computing and STEM settings predominate. Recent intervention studies reflect the same pattern, with training programs reported for pre-service physics teachers (Abdulayeva et al., 2025) and interventions that target awareness and attitudes rather than competence (Bilbao-Eraña & Arroyo-Sagasta, 2025), while comparable evidence for humanities and arts disciplines remains sparse. Reviews of AI in higher education report a similar distribution of research attention (Crompton & Burke, 2023; Zawacki-Richter et al., 2019; Zhai et al., 2021), and recommendations for AI in teacher education have been developed largely from these settings (García-Peñalvo, 2023). These reviews map how AI is used across higher education rather than how teachers are prepared to teach with it, so their evidence speaks to the overall distribution of research attention rather than to teacher competence specifically. Existing literature therefore points to a persistent gap concerning AI literacy among pre-service teachers outside technical or information-related disciplines. A second constraint concerns what has been measured. Much of this literature reports perceptions, interest, or educational needs (Dayagbil et al., 2025; Y. Kim & Choi, 2022; Moon et al., 2021), which indicate willingness to engage but not the competencies teachers would draw on when designing instruction. Analyses of teacher training programs reach a similar conclusion, identifying a shortage of research that addresses AI literacy across all teaching majors rather than within information education alone (Jeon et al., 2020). Studies that do address competence often report AI literacy as a single composite, leaving the internal structure of teachers’ preparation unexamined. Zhang et al. (2023) report moderate acceptance of AI in education among pre-service teachers without disaggregating by disciplinary background, and Gamlem et al. (2026) similarly find broad familiarity with generative AI alongside limited selfassessed expertise in a sample drawn across programs. It therefore remains unclear whether the moderate levels of acceptance and self-assessed expertise reported reflect a uniform profile or an average across uneven ones. This distinction is not incidental. A teacher whose competencies are moderate across all dimensions and a teacher whose ethical awareness is high while technical skill is low present different problems for curriculum design and targeted instructional support, yet both yield the same composite score.

151

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Competencies for AI Convergence Education Teacher competency has been conceptualized as an integrated set of personal characteristics, knowledge, skills, and attitudes required for effective performance in diverse teaching and learning contexts (Tigelaar et al., 2004). This integrative perspective has been widely adopted in teacher education research and provides a theoretical foundation for discussing emerging competencies related to AI convergence education. In the field of education, teacher competencies are increasingly framed around AI convergence education based on subject knowledge. UNESCO (2024) defines teachers’ AI competencies as the ability to acquire and actively apply skills necessary to solve academic or real-world problems using AI-related knowledge. In this study, AI convergence education competencies refer to the competencies teachers need to integrate AI into subject teaching, spanning foundational AI and pedagogical knowledge, the design and implementation of AI-based lessons, and the values that guide responsible use. Celik et al. (2022) and Park et al. (2021) state that foundational knowledge of AI and the ability to redesign curricula are essential competencies for AI convergence education. They particularly emphasize knowledge integration and curriculum reconstruction as key components. Heo and Kang (2023), drawing on a framework consistent with Celik (2023), divide teachers' AI convergence education competencies into two domains: (1) AI literacy, which includes understanding AI, practicing AI ethics, and problem-solving based on computational thinking; and (2) “AI application and convergence”, which includes AI-based lesson planning, development of instructional materials, lesson implementation, classroom management, and assessment. Frameworks for teachers’ AI convergence education competencies agree on two elements and diverge on a third. Foundational AI knowledge and the capacity to reconstruct curriculum appear across accounts (Celik et al., 2022; Lee & Kim, 2020; Park et al., 2021), as does the progression from lesson planning through implementation to assessment (Heo & Kang, 2023). Where the frameworks differ is in the status they assign to attitudes and values. Heo and Kang (2023) locate the practice of AI ethics within AI literacy, treating it as a knowledge-adjacent component, whereas UNESCO’s (2024) framework positions value orientation as a dimension distinct from knowledge and skills. This is more than a terminological choice. Situating ethics inside literacy implies that ethical practice follows from understanding AI, while separating the two implies that they may develop independently and at different rates. The available evidence does not settle the question, in part because most instruments report aggregated scores that cannot show whether the dimensions move together. Based on these prior studies, the present study analyzes the competencies of pre-service teachers majoring in general subjects in three dimensions, namely AI-related literacy, AI-based lesson design and implementation, and the value component of AI convergence education, a structure that allows the dimensions to be compared directly within the same sample.

152

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Theoretical Framework This study draws on two complementary theoretical foundations. First, the conceptualization of AI literacy proposed by Long and Magerko (2020) provides a multidimensional understanding of what it means to be AI-literate, encompassing not only technical comprehension but also critical evaluation and ethical engagement with AI systems. This framework has been widely adopted in educational research (Ng et al., 2021) and informs the operationalization of AI literacy across the cognitive, applied, and ethical dimensions examined in this study. Two qualifications apply to its use here. The framework was formulated to describe AI literacy for general audiences rather than for teachers, and it does not specify how understanding of AI translates into instructional decisions. It is therefore used to structure the measurement of AI literacy, while the pedagogical dimension is addressed through the competency framework described below. Second, the integrative conception of teacher competency articulated by Tigelaar et al. (2004), which defines competency as an integrated set of knowledge, skills, and attitudes enacted in professional contexts, provides the structural basis for understanding AI convergence education competencies. Within this framework, competency is not reducible to technical skill alone but encompasses pedagogical judgment, disciplinary knowledge, and value orientation, all of which are reflected in the three-component structure (foundational knowledge, implementation, and values) used to measure AI convergence education competencies in this study. Together, these frameworks position AI preparedness in teacher education as a multidimensional, context-sensitive construct rather than a uniform set of skills, aligning with the study's emphasis on disciplinary variation and the interplay between technical and valueoriented dimensions of AI competency. Methodology Research Design This study employed a quantitative, cross-sectional survey design. This design was appropriate for the research questions, which concern the levels of pre-service teachers' AI literacy and AI convergence education competencies (RQ1) and differences in these competencies across academic major, gender, level of AI understanding, and prior AI-related educational experience (RQ2 and RQ3). Because the study aimed to describe current competency levels and to compare independent subgroups rather than to establish causal relationships, a descriptive and comparative survey design using validated self-report instruments was adopted.

153

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Participants The participants of this study were pre-service teachers enrolled in the College of Education at a national teacher education institution in South Korea (hereafter University A), majoring in humanities, social sciences, or arts and physical education. Participants were recruited through convenience sampling at University A. A total of 112 students participated in the study. The general characteristics of the participants are presented in Table 1. By gender, there were 65 female students (58.04%) and 47 male students (41.96%). In terms of academic year, 57 students (50.89%) were in their first year, 32 (28.57%) in their second year, and 23 (20.54%) were third-year or higher. The participants' majors were classified as follows: Korean language education, social studies education, English education, early childhood education, and ethics education were categorized under the humanities and social sciences; physical education and music education were categorized under the arts and physical education. Among the participants, 74 students (66.07%) had prior experience with AI-related education, while 38 students (33.93%) had no such experience. Table 1 Characteristics of Participants

Gender

Major

Grade

Experience

Characteristics

Ratio (%)

Male

41.96

Female

58.04

Korean language

8.04

Social studies

8.93

English

17.86

Early childhood

19.64

Ethics

12.50

Physical

16.96

Music

16.07

First year

50.89

Second year

28.57

Third year or above

20.54

Experienced

66.07

Inexperienced

33.93

Instruments This study aimed to measure pre-service teachers’ AI literacy and AI convergence education competencies. To this end, prior studies were reviewed to select measurement tools appropriate

154

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

for pre-service teachers. Minor wording modifications were made to improve clarity and contextual appropriateness for pre-service teachers, without altering the original factor structure or the number of items in the instruments. The instruments used in this study included the AI Literacy Diagnostic Tool developed by Lim (2023), and the AI Convergence Competencies Scale developed by D. Kim et al. (2023), both designed for pre-service teachers. These instruments were selected because they were developed and validated within the context of Korean teacher education, reflecting the national curriculum and policy environment relevant to the participants of this study. Importantly, the conceptual domains of both instruments align closely with internationally recognized frameworks for AI literacy and teacher competency. The four-domain structure of the AI literacy diagnostic tool, encompassing AI understanding, application, development, and ethics, is consistent with the multidimensional conceptualization of AI literacy proposed by Long and Magerko (2020) and elaborated by Ng et al. (2021). Similarly, the three-component structure of the AI convergence competencies scale, comprising foundational knowledge, implementation, and values, reflects the integrative conception of teacher competency outlined by Tigelaar et al. (2004) and is further aligned with the knowledge, skills, and values dimensions articulated in UNESCO's (2024) AI Competency Framework for Teachers. This conceptual alignment supports the cross-contextual interpretability of the findings beyond the Korean educational context. The AI literacy diagnostic tool consists of 33 items across four domains (AI understanding, AI application, AI development, and AI ethics) and eight subdomains: basic AI knowledge, social impact of AI, AI technology use, problem-solving using AI, data literacy, basic programming, computational thinking, and AI ethics. The AI convergence competencies scale includes 34 items covering three components: foundational knowledge (AI knowledge, AI education knowledge, subject-convergent knowledge), implementation (AI convergence lesson design, operation, and support), and values (AI ethics, openness to AI, and AI teacher efficacy). The scale was designed specifically for pre-service teachers with limited classroom experience, emphasizing competencies such as value orientation and openness to AI convergence education rather than classroom management or assessment skills. All items on both instruments were rated on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree), with higher scores indicating higher self-assessed competence. In the present sample, internal consistency (Cronbach's α) was .94 for the AI literacy tool and .95 for the AI convergence competencies scale. Because these coefficients are based on 33 and 34 items respectively, reliability was also examined at the subdomain level, where α ranged from .74 to .89 across the eight subdomains of the AI literacy tool and from .76 to .90 across the nine subdomains of the AI convergence competencies scale, all exceeding the conventional criterion of .70. Both instruments had been validated previously in studies involving preservice teachers (D. Kim et al., 2023; Lim, 2023). To further ensure content validity in the present study, three experts in teacher education and educational technology reviewed the adapted items for theoretical consistency with the original instruments and alignment with the study's conceptual framework prior to data collection.

155

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Data Collection and Analysis Prior to the main study, a pilot survey was conducted over five days in August 2025 with three pre-service teachers from a comparable institution. Based on their feedback, minor wording adjustments were made to improve clarity, while the content, structure, and number of items remained unchanged. The main survey was administered online using Google Forms to preservice teachers at University A. Data collection took place over three weeks, from September 5 to September 26, 2025. A total of 121 responses were collected, and 112 responses were retained after excluding incomplete or careless responses. The data were analyzed using SPSS Statistics version 29. Descriptive statistics and frequency analysis were conducted to examine the participants’ background variables and to calculate the mean and standard deviation of each item. To assess the competency levels of pre-service teachers, one-sample t-tests were conducted using 3.00 as the test value, representing the scale midpoint of the 5-point Likert scale. In addition, independent samples t-tests and one-way ANOVA were performed to examine mean differences based on major type (humanities/social science vs. arts/physical education), gender, and self-perceived level of AI understanding. Participants' levels of AI understanding were categorized into three groups (high, medium, and low) based on the sample mean and standard deviation to facilitate comparative analysis. Because the analysis involved multiple comparisons across subdomains within each research question, the false discovery rate was controlled using the Benjamini-Hochberg procedure. The false discovery rate rather than the family-wise error rate was controlled because the subgroup comparisons in this study were exploratory. The correction was applied separately to each family of tests reported in a given table, and the outcome of the adjustment is reported in the Results. Effect sizes were reported so that the magnitude of each difference could be evaluated independently of significance testing, using Cohen’s d for two-group comparisons and eta squared for analysis of variance. The selection of grouping variables for each dependent variable was guided by theoretical relevance and prior empirical evidence. For AI literacy, academic major, gender, and level of AI understanding were examined as grouping variables, as these factors have been identified in previous research as meaningful sources of variation in AI-related knowledge and attitudes among pre-service teachers (Gamlem et al., 2026; Laru et al., 2025; Park, 2021). For AI convergence education competencies, prior AI-related educational experience was selected as the primary grouping variable, given its direct relevance to the acquisition of pedagogical competencies for AI integration (Abdulayeva et al., 2025; Jeon et al., 2020). Major groupings for comparative analysis were consolidated into two categories, humanities and social sciences versus arts and physical education, because the limited sample size within individual majors would have compromised the statistical power of subgroup analyses. Ethical Considerations Participation in this study was voluntary, and informed consent was obtained from all participants prior to data collection. Participants were informed of the purpose of the study, the anonymous nature of the survey, and their right to withdraw at any time without penalty. No

156

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

personally identifiable information was collected, and all data were analyzed in aggregate form. The study involved a non-invasive, minimal-risk survey of adult pre-service teachers and did not include any intervention or manipulation. Accordingly, the study was conducted in accordance with commonly accepted ethical practices in educational research, including respect for persons, confidentiality, and data protection. Results Analysis of Pre-service Teachers’ AI Literacy Results of AI Literacy by Domain This subsection presents the descriptive statistics and one-sample t-test results for each domain and subdomain of AI literacy. In addition to statistical significance testing, effect sizes (Cohen’s d) were calculated for the one-sample t-test results to examine the practical magnitude of the observed differences. The overall mean score for AI literacy among pre-service teachers was 3.32 (SD = 0.48). Domain-level analysis results are presented in Table 2. Overall, participants scored significantly above the scale midpoint of 3.00 across most subdomains of AI literacy. In the AI Understanding domain, both basic knowledge and social impact showed statistically significant positive deviations from the scale midpoint (p < .01). Within the AI Application domain, technology utilization and problem-solving also yielded significantly higher scores (p < .01), whereas data literacy did not differ significantly from the scale midpoint (p = .516). Unlike other subdomains that significantly exceeded the midpoint, data literacy remained at the scale midpoint, suggesting a comparatively less developed area rather than a significantly deficient one. In the AI Development domain, computational thinking was significantly higher than the scale midpoint (p < .01), while programming scores were significantly lower (p < .01). In addition, participants demonstrated significantly higher scores in the AI Ethics subdomain (p < .01).

157

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Table 2 Descriptive Statistics and One-Sample t-Test Results for Pre-Service Teachers' AI Literacy Factor AI Understanding AI Application

AI Development

Cohen’s d **

basic knowledge

112

3.23

.704

3.48

.001

0.33

social impact

112

3.52

.646

8.59

.001**

0.81

technology utilization

112

3.93

.752

13.1 9

.001**

problem solving

112

3.46

.668

7.31

.001**

0.69

data literacy

112

2.95

.774

-.65

.516

-0.06

programming

112

2.40

.723

-8.72

.001**

-0.82

computational thinking

112

3.35

.689

5.48

.001**

112

3.73

.738

10.5 8

.001**

AI Ethics

1.25

0.52 1.00

Note. The test value (3.00) represents the neutral midpoint of the 5-point Likert scale. *p<.05, **p<.01

Analysis by Participant Characteristics This subsection reports differences in AI literacy according to participant characteristics, including major, gender, and level of AI understanding. Analysis by major revealed statistically significant differences in selected subdomains of AI literacy (Table 3). Pre-service teachers majoring in humanities and social sciences scored significantly higher than those majoring in arts and physical education in the social impact and AI ethics subdomains (p < .01), with medium effect sizes (d = 0.65 and d = 0.66, respectively). No significant differences were observed between the two groups in the AI application or AI development domains, and effect sizes for these subdomains were small or negligible (d = 0.11–0.38). Table 3 Independent Samples t-Test Results for AI Literacy Subdomains by Academic Major Factor basic knowledge AI Understanding

AI Application

social impact technology utilization

Major

3.29

.653

3.13

.798

3.65

.550

3.25

.743

4.03

.630

3.74

.933

158

Cohen's d

1.025

.308

0.21

3.246

.002**

0.65

1.898

.060

0.38

IAFOR Journal of Education: Technology in Education

problem Solving data literacy programming AI Development computational thinking

AI Ethics

Volume 14 – Issue 2– 2026

3.51

.561

3.35

.844

3.00

.734

2.84

.852

2.43

.730

2.34

.715

3.38

.650

3.30

.768

3.89

.648

3.42

.815

1.079

.286

0.22

1.008

.316

0.20

.590

.556

0.12

.571

.569

0.11

3.304

.001**

0.66

Note. Major 1 = Humanities and Social Sciences (n = 75); Major 2 = Arts and Physical Education (n = 37). *p<.05, **p<.01

Gender-based comparisons indicated that female participants scored significantly higher than male participants in basic knowledge and social impact within the AI Understanding domain (p < .01), with medium effect sizes (d = 0.61 and d = 0.62, respectively). In the AI Application domain, female participants also demonstrated higher scores in technology utilization (p < .01, d = 0.54) and problem-solving (p < .05, d = 0.50), both reflecting medium effect sizes. In contrast, no statistically significant gender differences were found in data literacy, basic programming, or computational thinking, and effect sizes for these subdomains were small or negligible (d = 0.06–0.29). In the AI Ethics domain, female participants scored significantly higher than male participants (p < .05), with a small-to-medium effect size (d = 0.45). Table 4 Independent Samples t-Test Results for AI Literacy Subdomains by Gender Factor

basic knowledge AI Understanding

Gender

male

4 7

2.99

.796

female

6 5

3.40

.576

male

4 7

3.28

.807

female

6 5

3.70

.425

male

4 7

3.70

.861

female

6 5

4.10

social impact

AI Application

technology utilization

159

.617

Cohen's d

-3.179

.002**

0.61

-3.260

.002**

0.62

-2.815

.006**

0.54

IAFOR Journal of Education: Technology in Education

problem Solving

Volume 14 – Issue 2– 2026

male

4 7

3.27

female

6 5

3.60

.553

male

4 7

2.85

.883

female

6 5

3.02

.683

male

4 7

2.28

.746

female

6 5

2.49

.698

male

4 7

3.38

.833

female

6 5

3.33

.568

male

4 7

3.54

.849

female

6 5

3.87

data literacy

programming AI Development computational thinking

AI Ethics

.767 -2.635

.010*

0.50

-1.133

.260

0.22

-1.536

.127

0.29

.317

.752

0.06

-2.343

.021*

0.45

.617

Note. *p<.05, **p<.01

Differences in AI literacy according to participants’ level of AI understanding are presented in Table 5. Significant group differences were observed in six subdomains: basic knowledge, social impact, problem-solving, data literacy, computational thinking, and AI ethics. Effect sizes varied considerably across subdomains. The largest effect was observed in basic knowledge (η² = .701), indicating that perceived AI understanding level accounted for a substantial proportion of variance in this subdomain. Large effects were also found in social impact (η² = .166) and AI ethics (η² = .163), while problem-solving (η² = .112) and computational thinking (η² = .116) showed medium effects, and data literacy showed a medium-to-large effect (η² = .137). No significant differences were found among groups in technology utilization or programming ability, and effect sizes for these subdomains were small (η² = .034–.046). These results indicate that higher perceived AI understanding is associated with broader strengths across multiple dimensions of AI literacy, with particularly pronounced differences in foundational knowledge.

160

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Table 5 Comparison of AI Literacy Subdomains Across Levels of AI Understanding Factor

AI Understanding

Level basic knowledge social impact technology utilization

AI Application

problem solving

data literacy

AI Development

programmin g computation al thinking

M higha b

medium

lowc higha mediumb lowc higha mediumb lowc higha mediumb lowc higha mediumb lowc higha mediumb lowc higha mediumb lowc 4.21

SD 4.14

F(p) .273

3.46

.168

2.69 3.86 3.71 3.26 4.18 4.00 3.79 3.76 3.61 3.24 3.45 2.95 2.72 2.64 2.37 2.31 3.73 3.43 3.14 .484

higha AI Ethics

.501 .427 .313 .755 .577 .516 .893 .551 .515 .723 .599 .660 .806 .676 .792 .690 .572 .608 .708 10.59 0 (<.001

Scheffe

η²

127.566 (<.001***

c<a

0.701

10.855 (.001**)

b,c<a

0.166

2.607 (.078)

0.046

6.859 (.003**)

c<a

0.112

8.664 (.002**)

b,c<a

0.137

1.947 (.148)

0.034

7.173 (.006**)

c<a

0.116

c<a

0.163

)

***)

mediumb lowc

3.83 3.47

.574 .798

Note. Means with different superscripts differ significantly at p < .05 based on Scheffé post hoc comparisons. *p<.05, **p<.01, ***p<.001

Analysis of Pre-service Teachers’ AI Convergence Competencies Results of AI Convergence Education Competencies by Domain This subsection reports domain- and subdomain-level results of AI convergence education competencies based on one-sample t-tests. The overall mean score for AI convergence

161

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

education competencies was 3.49 (SD = 0.67). The domain-level analysis results are presented in Table 6. In the foundational knowledge component, statistically significant differences from the scale midpoint of 3.00 were observed for AI education knowledge and subject-convergent knowledge (p < .01), whereas basic AI knowledge did not show a significant difference. In the AI Application component, all three subdomains, namely lesson design, implementation, and support, scored significantly above the scale midpoint (p < .01). Similarly, all subdomains within the AI Values component, including ethics, openness, and teacher efficacy, demonstrated significantly higher scores than the scale midpoint (p < .01). Table 6 Descriptive Statistics and One-Sample t-Test Results for Pre-Service Teachers' AI Convergence Education Competencies

AI Knowledge

Factor

Cohen’s d

basic

112

3.06

.704

.97

.333

0.09

education

112

3.34

.710

5.12

.001**

0.48

subject-convergence

112

3.21

.769

3.00

.003**

0.28

designing

112

3.39

.623

6.69

.001**

0.63

112

3.20

.702

3.01

.003**

0.28

AI implementing Application supporting AI Values

112

3.44

.573

8.15

.001

0.77

ethics

112

4.05

.552

20.27

.001**

1.92

openness

112

3.91

.648

14.86

.001**

1.40

11.15

1.05

teacher efficacy

112

3.73

.700

.001

Note. The test value (3.00) represents the neutral midpoint of the 5-point Likert scale. *p<.05, **p<.01

Analysis by Participant Characteristics This subsection examines differences in AI convergence education competencies according to participants’ prior experience with AI-related education (Table 7). Significant differences were found between participants with and without AI-related educational experience across all three subdomains of the foundational knowledge component. The difference in basic AI knowledge was particularly pronounced, yielding a very large effect size (d = 2.31, p < .001), suggesting that prior AI educational experience is strongly associated with foundational knowledge acquisition. Medium effect sizes were observed for subjectconvergent knowledge (d = 0.77, p < .001) and AI education knowledge (d = 0.58, p < .01). Within the AI Application component, significant group differences were observed in implementation (d = 0.60, p < .01) and support (d = 0.64, p < .01), both reflecting medium effect sizes, whereas no significant difference was found in lesson design (d = 0.37, p = .069). In the AI Values component, teacher efficacy differed significantly between groups (d = 0.50, p < .05), with higher scores among participants with prior AI-related educational experience.

162

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

In contrast, no significant differences were observed in ethics or openness, and effect sizes for these subdomains were small (d = 0.29–0.37), indicating that these value-oriented dimensions may be less directly influenced by prior AI education experience. Table 7 Independent Samples t-Test Results for AI Convergence Education Competencies by AI Education Experience Factor basic AI Knowledge

education subjectconvergence designing

AI Application

implementing supporting ethics

AI Values

openness teacher efficacy

Experience

3.43

.404

2.23

.519

3.47

.662

3.07

.728

3.42

.596

2.80

.893

3.46

.630

3.24

.565

3.33

.615

2.95

.773

3.55

.524

3.20

.584

4.11

.542

3.95

.567

4.00

.627

3.75

.674

3.87

.543

3.47

.889

Cohen’ sd

11.59 <.001***

2.31

2.92

.004**

0.58

3.85

<.001***

0.77

1.83

.069

0.37

3.03

.003**

0.60

3.19

.002**

0.64

1.47

.144

0.29

1.87

.064

0.37

2.53

.015*

0.50

Note. Experience 1 = Experienced (n = 74); Experience 2 = Inexperienced (n = 38). *p<.05, **p<.01, ***p<.001

As described in the Methodology, the false discovery rate was controlled across the comparisons reported within each table. The adjustment did not alter the significance status of any comparison reported in Tables 2 through 7 at the .05 level. All differences identified as significant remained significant after adjustment, and all non-significant differences remained non-significant.

163

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Discussion Summary of Key Findings This study examined pre-service teachers’ AI literacy and AI convergence education competencies, focusing on differences across domains and participant characteristics. Overall, participants demonstrated levels above the scale midpoint in most domains of both AI literacy and AI convergence education competencies, suggesting a generally positive baseline of AIrelated preparedness among pre-service teachers. However, while most subdomains exceeded the scale midpoint, data literacy remained at the scale midpoint and basic programming skills were significantly lower, indicating uneven development across subdomains. In addition, meaningful differences emerged according to academic major, gender, level of AI understanding, and experience with AI-related education. These findings suggest that preservice teachers’ preparedness for AI-integrated education is not uniform but shaped by a combination of disciplinary background and individual learning experiences. Furthermore, participants with higher levels of AI understanding and prior experience with AIrelated education demonstrated stronger competencies in most, though not all, subdomains. The exceptions were technology utilization and programming for level of AI understanding, and lesson design, ethics, and openness for prior educational experience. This finding underscores the cumulative nature of AI preparedness, suggesting that early and sustained exposure to AI-related learning opportunities may play a critical role in developing both technical and pedagogical competencies. These patterns bear on a question the existing frameworks leave open. Accounts of teachers’ AI convergence education competencies differ in whether value orientation is treated as a component of AI literacy or as a dimension that develops alongside it, and the aggregated scores typically reported in prior studies cannot distinguish the two positions. The present findings show ethical awareness and openness at their highest levels in the same sample where data literacy sits at the scale midpoint and programming falls below it. Value-oriented and technical dimensions did not move together here, which is consistent with treating them as separable rather than as facets of a single construct. This observation is drawn from a single cross-sectional sample and describes a pattern rather than a developmental sequence, so it should be read as a case for measuring the dimensions separately rather than as evidence that one precedes the other. Technical Gaps in AI Literacy and AI Convergence Education Data literacy did not significantly exceed the scale midpoint, unlike most other subdomains, and programming-related competencies were significantly lower. This pattern suggests that pre-service teachers may have limited opportunities to engage with data-driven practices and AI development activities within existing teacher education programs. However, this pattern warrants deeper consideration beyond a simple lack of technical exposure. In many teacher

164

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

education curricula, AI-related content tends to be positioned as a conceptual topic that emphasizes definitions, social implications, or ethical concerns, rather than as a methodological tool embedded in instructional practice. From this perspective, the observed gap in data literacy and programming competencies may reflect a structural separation between pedagogical training and technical skill development in pre-service teacher education. As a result, pre-service teachers may develop an awareness of AI and its implications without acquiring the analytical competencies needed to meaningfully engage with data-driven instructional tools. This imbalance is particularly concerning in teacher education contexts, where teachers are increasingly expected to interpret learning analytics, evaluate AI-based assessment tools, and make instructional decisions informed by student data. When data-related competencies remain at the scale midpoint and programming-related competencies are underdeveloped, future teachers risk becoming passive users of AI-enabled technologies rather than critical evaluators of their pedagogical affordances and limitations. Such a tendency may limit teachers’ ability to adapt AI tools to diverse classroom contexts and could inadvertently reinforce existing digital divides, particularly for students who rely on teachers to mediate access to and understanding of AI-supported learning environments. Accordingly, strengthening technical dimensions of AI literacy should be viewed not as an optional enhancement but as a core component of equitable and responsible AI integration in teacher education. These findings are consistent with Long and Magerko’s (2020) conceptualization of AI literacy as a multidimensional construct in which technical execution, including data literacy and programming, is as foundational as conceptual awareness. The gap observed in these subdomains suggests that current teacher education programs may be fostering the awareness dimension of AI literacy without sufficiently developing its technical dimension. Disciplinary and Individual Differences in AI Preparedness Differences in AI literacy across academic majors suggest that disciplinary backgrounds influence how pre-service teachers interpret and engage with AI-related issues. Pre-service teachers in the humanities and social sciences demonstrated higher awareness of the social impact and ethical dimensions of AI, reflecting the epistemological orientations of these disciplines toward societal and value-based inquiry. Gender-based differences were also observed in several subdomains of AI literacy, including AI understanding, application, and ethics. While these differences should be interpreted cautiously, they point to the need for teacher education programs to consider how instructional approaches may differentially support diverse learner groups. Gender differences in attitudes towards generative AI have also been reported in other teacher education contexts (Gamlem et al., 2026). These findings may reflect broader patterns in digital technology engagement, where female pre-service teachers in non-STEM disciplines may be more actively exposed to AI tools

165

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

in everyday contexts, or may demonstrate greater motivation to engage with AI in pedagogically oriented ways. The present data cannot distinguish between these accounts. Taken together, these profiles qualify rather than simply confirm Tigelaar et al.’s (2004) integrative conception of teacher competency. That account holds that knowledge, skills, and attitudes constitute an integrated whole, and it is often invoked to justify treating competency as a single construct. The present results are consistent with the integrative claim in the sense that all three elements are present, but they also show the elements at markedly different levels within the same individuals, with value-oriented dispositions toward AI developing ahead of the technical competencies needed to enact them in practice. Integration, on this evidence, describes what competency requires rather than what teacher preparation currently produces. For measurement, the implication is that composite competency scores may obscure the differences that matter most for curriculum design. Recommendations The findings of this study point to several recommendations for the design and implementation of teacher education programs in the context of AI convergence education. First, teacher education programs should move beyond a one-size-fits-all approach and adopt discipline-sensitive instructional designs that align AI competencies with subject-specific ways of knowing and teaching. The significant differences observed across academic majors, together with the low programming scores and the absence of any significant elevation in data literacy across groups, suggest that current introductory AI courses may not be sufficiently differentiated to address the diverse learning needs of pre-service teachers. Disciplineembedded approaches may allow technical competencies to develop in tandem with existing disciplinary strengths, thereby reducing barriers to engagement (Chiu & Chai, 2020). Examples include Natural Language Processing-based text analysis for humanities and social science majors, and generative AI tools for creative and movement-based applications in arts and physical education. Second, given the very large effect size associated with prior AI-related educational experience on foundational knowledge acquisition (d = 2.31), teacher education institutions should prioritize early and sustained exposure to AI-related learning opportunities. The substantial gap between experienced and inexperienced participants suggests that a single introductory course may be insufficient; instead, sequenced and cumulative curricular experiences that build AI competencies progressively across the degree program are warranted. Third, while ethics and openness within AI convergence education competencies scored well above the midpoint (M = 4.05 and M = 3.91, respectively; Table 6) and did not differ significantly by prior AI-related educational experience (Table 7), these should not be treated as sufficient substitutes for technical competency. Programs should ensure that ethical and dispositional strengths are complemented by structured opportunities to develop data literacy

166

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

and applied AI skills, so that pre-service teachers are equipped to critically evaluate, rather than passively adopt, AI-enabled educational tools. Finally, given the observed gender differences in AI literacy across multiple subdomains, with medium effect sizes across AI understanding, application, and ethics (d =0.45–0.62; Table 4), teacher education programs should examine whether current instructional approaches inadvertently disadvantage particular learner groups. Inclusive pedagogical strategies that make AI-related content accessible regardless of prior technical background may help ensure more equitable outcomes in AI preparedness across diverse pre-service teacher populations. Conclusion This study investigated AI literacy and AI convergence education competencies among preservice teachers across diverse academic majors, addressing a gap in the existing literature that has primarily focused on elementary education or technology-oriented teacher education contexts. By examining pre-service teacher preparation within predominantly non-STEM subject areas, this study extends current understanding of how future teachers are being prepared for AI-integrated education. The study contributes on two levels. Empirically, it provides subdomain-level evidence on a population that prior research has largely passed over, namely pre-service teachers preparing to teach non-STEM subjects, and it confirms that AI preparedness in this group is multidimensional, encompassing technical knowledge, pedagogical application, and ethical orientation within subject-specific contexts. Conceptually, it shows that the dimensions assumed to be integrated in existing competency frameworks can be observed at substantially different levels within the same individuals, which supports reporting these dimensions separately rather than as a composite. Read together, the two contributions indicate that questions about whether teachers are prepared for AI-integrated education are answered poorly by a single index, and better by a context-sensitive profile that shows where preparation is strong and where it is absent. Despite its contributions, this study has several limitations. First, the use of convenience sampling from a single teacher education institution limits the generalizability of the findings. In addition, although the false discovery rate was controlled across families of comparisons, the number of subgroup analyses relative to the sample size means that these findings are best understood as patterns requiring replication rather than as definitive effects. Institutional curriculum structures and contextual factors may have influenced participants’ AI literacy and convergence competencies; therefore, the results should be interpreted as context-specific rather than representative of all pre-service teachers in South Korea. Second, this study relied on self-reported survey data, which may not fully capture participants’ actual technical competencies or instructional practices. Future research could incorporate performance-based assessments or scenario-based tasks to examine how pre-service teachers apply AI-related knowledge in instructional contexts. The validity evidence for the adapted instruments is also

167

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

limited. Because the sample of 112 was not large enough to support a factor analysis, the factor structure of the instruments was not re-examined in the present study and requires confirmation in a larger and independent sample. Finally, the cross-sectional design limits conclusions about developmental changes over time. Longitudinal and multi-institutional studies would provide deeper insight into how AI literacy and convergence competencies evolve across different stages and contexts of teacher education.

Declaration of Generative AI and AI-assisted Technologies The author would like to acknowledge the use of artificial intelligence (AI) tools in the preparation of this manuscript. The following tool was used for this specific purpose only: Google Gemini was used to translate the manuscript from Korean into English. All AItranslated text was thoroughly reviewed and revised by the author to ensure accuracy, clarity, and adherence to the intended meaning of the original text. AI tools were not used for other purposes, including data generation, data analysis, methodology, or interpretation of findings or conclusions. The author takes full responsibility for all aspects of the final manuscript.

168

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

References Abdulayeva, A., Zhanatbekova, N., Andasbayev, Y., & Boribekova, F. (2025). Fostering AI literacy in pre-service physics teachers: Inputs from training and co-variables. Frontiers in Education, 10, 1505420. https://doi.org/10.3389/feduc.2025.1505420 Bilbao-Eraña, A., & Arroyo-Sagasta, A. (2025). Fostering AI literacy in pre-service teachers: Impact of a training intervention on awareness, attitude and trust in AI. Frontiers in Education, 10, 1668078. https://doi.org/10.3389/feduc.2025.1668078 Casal-Otero, L., Catala, A., Fernández-Morante, C., Taboada, M., Cebreiro, B., & Barro, S. (2023). AI literacy in K-12: A systematic literature review. International Journal of STEM Education, 10(1), 29. https://doi.org/10.1186/s40594-023-00418-7 Celik, I. (2023). Towards intelligent-TPACK: An empirical study on teachers' professional knowledge to ethically integrate artificial intelligence (AI)-based tools into education. Computers in Human Behavior, 138, 107468. https://doi.org/10.1016/j.chb.2022.107468 Celik, I., Dindar, M., Muukkonen, H., & Järvelä, S. (2022). The promises and challenges of artificial intelligence for teachers: A systematic review of research. TechTrends, 66(4), 616–630. https://doi.org/10.1007/s11528-022-00715-y Chiu, T. K. F., & Chai, C. S. (2020). Sustainable curriculum planning for artificial intelligence education: A self-determination theory perspective. Sustainability, 12(14), Article 5568. https://doi.org/10.3390/su12145568 Crompton, H., & Burke, D. (2023). Artificial intelligence in higher education: The state of the field. International Journal of Educational Technology in Higher Education, 20(1), 22. https://doi.org/10.1186/s41239-023-00392-8 Dayagbil, F. T., Boholano, H. B., & Sumalinog, G. G. (2025). Are they in or out? Exploring pre-service teachers’ knowledge, perceptions, and experiences regarding artificial intelligence (AI) in teaching and learning. Frontiers in Education, 10, 1665205. https://doi.org/10.3389/feduc.2025.1665205 Gamlem, S. M., McGrane, J., Brandmo, C., Moltudal, S., Sun, S. Z., & Hopfenbeck, T. N. (2026). Exploring pre-service teachers’ attitudes and experiences with generative AI: A mixed methods study in Norwegian teacher education. Educational Psychology, 46(1), 27-51. https://doi.org/10.1080/01443410.2025.2528663 García-Peñalvo, F. J. (2023). La percepción de la Inteligencia Artificial en contextos educativos tras el lanzamiento de ChatGPT: Disrupción o pánico [The perception of artificial intelligence in educational contexts after the launch of ChatGPT: Disruption or panic?]. Education in the Knowledge Society, 24, Article e31279. https://doi.org/10.14201/eks.31279 Heo, H., & Kang, S. (2023). Teacher competencies for designing artificial intelligenceintegrated education. Journal of Korean Association for Computer Education, 26(2), 89-100. https://doi.org/10.32431/kace.2023.26.2.008 Jeon, I., Jeon, S., & Song, K. (2020). Teacher training programs and teacher needs analysis to strengthen AI education competencies. Journal of Korean Association for Information Education, 24(4), 279-289. https://doi.org/10.14352/jkaie.2020.24.4.279

169

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Kim, D., So, H., & Lim, J. Y. (2023). Development the measurement instrument of AI convergence education competency for pre-service teachers. The Journal of Educational Studies, 54(3), 139-168. https://doi.org/10.15854/jes.2023.09.54.3.139 Kim, N., & Kim, M. (2024). The current status, perceptions, and demands for AI education among prospective early childhood teachers. Korean Journal of Early Childhood Education, 26(1), 147-168. https://doi.org/10.15409/riece.2024.26.1.7 Kim, Y., & Choi, H. (2022). Kindergarten teachers’ perceptions of AI education for early childhood. Learner-Centered Curriculum Studies, 22(6), 163178. https://doi.org/10.22251/jlcci.2022.22.6.163 Laru, J., Celik, I., Jokela, I., & Mäkitalo, K. (2025). The antecedents of pre-service teachers’ AI literacy: Perceptions about own AI driven applications, attitude towards AI and knowledge in machine learning. European Journal of Teacher Education, 48(5), 964986. https://doi.org/10.1080/02619768.2025.2535623 Lee, S. (2020). Elementary teachers' understanding and awareness of AI education. Korean Journal of Elementary Education, 31, 15-31. https://doi.org/10.20972/kjee.31.202008.15 Lee, W., & Kim, J. (2020). Curriculum development for AI convergence education. Korean Journal of Converging Humanities, 8(3), 29-52. https://doi.org/10.14729/converging.k.2020.8.3.29 Lim, H. (2023). Development and validation of AI literacy diagnostic tool for prospective secondary teachers (Doctoral dissertation). Gyeongsang National University. Long, D., & Magerko, B. (2020). What is AI literacy? Competencies and design considerations. In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems (pp. 1-13). https://doi.org/10.1145/3313831.3376727 Maslej, N., Fattorini, L., Perrault, R., Gil, Y., Parli, V., Kariuki, N., Capstick, E., Reuel, A., Brynjolfsson, E., Etchemendy, J., Ligett, K., Lyons, T., Manyika, J., Niebles, J. C., Shoham, Y., Wald, R., Hamrah, A., Santarlasci, L., Lotufo, J. B., Rome, A., Shi, A., & Oak, S. (2025). The AI Index 2025 Annual Report. Stanford: AI Index Steering Committee, Institute for Human-Centered AI, Stanford University. Ministry of Education. (2023). Digital-based educational innovation plan. Sejong: Ministry of Education. Moon, K., Yang, J., & Park, S. (2021). University freshmen’s perceptions and directions for AI education as a liberal arts course. Korean Journal of General Education, 15(5), 1123. https://doi.org/10.46392/kjge.2021.15.5.11 Ng, D. T. K., Leung, J. K. L., Chu, K. W. S., & Qiao, M. S. (2021). AI literacy: Definition, teaching, evaluation and ethical issues. Proceedings of the Association for Information Science and Technology, 58(1), 504–509. https://doi.org/10.1002/pra2.487 OECD. (2019). OECD future of education and skills 2030: OECD learning compass 2030. OECD Publishing. Park, H., Kim, J., & Lee, W. (2021). Deriving teacher competencies for AI converged education. Journal of Korean Association for Computer Education, 24(5), 1725. https://doi.org/10.32431/kace.2021.24.5.002

170

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Park, S. (2021). Prospective teachers’ perceptions of AI education according to AI learning experience, interest, and major department. Journal of Korean Association for Information Education, 25(1), 103-111. https://doi.org/10.14352/jkaie.2021.25.1.103 Ryu, M., & Han, S. (2018). Elementary teachers’ educational perceptions of artificial intelligence. Journal of Korean Association for Information Education, 22(3), 317324. https://doi.org/10.14352/jkaie.2018.22.3.317 Tigelaar, D. E., Dolmans, D. H., Wolfhagen, I. H., & Van Der Vleuten, C. P. (2004). The development and validation of a framework for teaching competencies in higher education. Higher Education, 48(2), 253-268. https://doi.org/10.1023/B:HIGH.0000034318.74275.e4 Tondeur, J., van Braak, J., Sang, G., Voogt, J., Fisser, P., & Ottenbreit-Leftwich, A. (2012). Preparing pre-service teachers to integrate technology in education: A synthesis of qualitative evidence. Computers & Education, 59(1), 134–144. https://doi.org/10.1016/j.compedu.2011.10.009 UNESCO. (2019). Artificial intelligence in education: Promises and implications for teaching and learning. UNESCO Publishing. https://unesdoc.unesco.org/ark:/48223/pf0000366994 UNESCO. (2024). AI competency framework for teachers. UNESCO Publishing. https://doi.org/10.54675/ZJTE2084 Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of research on artificial intelligence applications in higher education: where are the educators? International Journal of Educational Technology in Higher Education, 16(1), 1–27. https://doi.org/10.1186/s41239-019-0171-0 Zhai, X., Wang, M., & Ghani, U. (2021). A systematic review of artificial intelligence applications in education from 2010 to 2020. Education and Information Technologies, 26, 6539–6560. https://doi.org/10.1007/s10639-021-10551-7 Zhang, C., Schießl, J., Plößl, L., Hofmann, F., & Gläser-Zikuda, M. (2023). Acceptance of artificial intelligence among pre-service teachers: A multigroup analysis. International Journal of Educational Technology in Higher Education, 20(1), Article 49. https://doi.org/10.1186/s41239-023-00420-7 Corresponding Author: Juyoung Lee Email: leejysam@daum.net

171

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

172

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Gamifying Sepedi Riddles for Developing Critical Thinking in Grade 7 Learners Nkame Emmanuel Ngobeni University of Limpopo, South Africa Ablonia Dihloriso Maledu University of Limpopo, South Africa

173

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Abstract Sepedi dithai (riddles) are a cognitively demanding sub-genre of indigenous folklore, yet their pedagogical potential remains largely confined to oral and print modalities. This study reports a design-based research study in which ten Sepedi dithai were integrated into a browser-based educational game and deployed to learners through Netlify. The study was guided by Vygotsky's sociocultural theory, Anderson and Krathwohl’s revised taxonomy of educational objectives, Plass, Homer, and Kinzer's foundations of game-based learning, and Mishra and Koehler's Technological Pedagogical Content Knowledge framework. A three-iteration design-based research cycle was conducted with four Sepedi home language teachers and twenty Grade 7 learners. Data were generated through non-participant classroom observations, semi-structured post-play interviews, game-log analytics, and document analysis of learners' typed in-game responses, and were analysed thematically. Findings indicated that the game sustained engagement with higher-order reasoning across inferential, evaluative, and creative riddle categories, stimulated culturally grounded reasoning in which learners drew on the indigenous concept of botho (humanness), extended teacher scaffolding into asynchronous modes, and strengthened the home-school connection through multigenerational play. The article proposes that carefully designed, culturally situated educational technology can amplify rather than displace indigenous folklore pedagogies and offers a potential replicable design pattern for other indigenous African languages. Keywords: critical thinking, digital game-based learning, design-based research, indigenous knowledge systems, Sepedi dithai

174

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Critical thinking is now widely treated as a non-negotiable outcome of 21st-century schooling (Alsaleh, 2020; Thornhill-Miller et al., 2023; Trilling & Fadel, 2009). In South African indigenous-language classrooms, learners continue to struggle with comprehension tasks that require analysis, inference, and evaluation (Howie et al., 2017; Pretorius & Spaull, 2022). A growing body of South African and continental studies argues that Sepedi and other indigenous-language folklore, when deployed with intentional scaffolding and higher-order questioning, can support the development of critical thinking in home language learners (Jaxa, 2024; Maledu, 2024; Wiysahnyuy & Anjirbag, 2023). Folklore-based pedagogies depend on rich, sustained teacher mediation that is difficult to maintain under typical large-class, resourceconstrained conditions. This study also reports on a technology-integrated approach to addressing this tension. Dithai (riddles) were selected as the focal folklore sub-genre for three reasons. Firstly, dithai are compact and self-contained, which makes them well-suited to short-cycle digital interaction (Finnegan, 2012; Okpewho, 1992). Secondly, by design, dithai require inference, evaluation, and creation, which sit at the higher levels of the revised taxonomy of educational objectives and are widely recognised as cognitively demanding for Grade 7 learners (Anderson & Krathwohl, 2001). Lastly, dithai remain a living part of everyday Sepedi cultural life, so their digital remediation offers a culturally responsive entry point for educational technology in Home Language classrooms. The Dithai Digital Game described in this article is a browser-based educational game implemented in HyperText Markup Language, Cascading Style Sheets, and JavaScript, and deployed as a progressive web application through Netlify (a cloud platform that hosts websites and web applications and makes them available to users over the internet). Netlify was selected because it provides continuous deployment from a Git repository, free HTTPS hosting, a global content delivery network, and offline-capable delivery to any device with a modern browser. These characteristics are well-suited to the irregular connectivity typical of rural Limpopo schools (Shaji, 2023). The game presents a bank of ten Sepedi dithai across four ascending cognitive levels derived from the revised taxonomy of educational objectives (Anderson & Krathwohl, 2001). The study asks: How can Sepedi dithai be integrated into a digital game to develop critical thinking in Grade 7 home language learners, and what cognitive, cultural, and pedagogical effects emerge when learners and teachers engage with such a game? In answering this question, the article contributes to three scholarly conversations: the conversation on criticalthinking pedagogy in indigenous African languages, the conversation on culturally responsive educational technology, and the conversation on the digitisation of folklore as an epistemic project. The primary focus of the study is the development of critical thinking; the riddles and the digital game are the means through which that outcome is pursued, not separate objects of study. The contribution of the study is threefold. Empirically, it documents what happens when ten authentic Sepedi dithai are translated into a gamified digital format and played by Grade 7 learners and their teachers in a specific rural South African circuit. Theoretically, it argues that 175

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

the sociocultural, ecological, and game-based learning frameworks can be productively combined to account for the cognitive, cultural, and pedagogical effects of digitised folklore. Practically, it offers a replicable design pattern, grounded in the revised taxonomy of educational objectives and in the Technological Pedagogical Content Knowledge (TPACK) framework (Mishra & Koehler, 2006), that other educators working with indigenous African languages can adapt for their own contexts. The remainder of the article is organised into a literature review, a theoretical framework, a method section, a findings section, a discussion, and a conclusion. Literature Review The Critical-Thinking Potential of Riddles Riddles are one of the most cognitively demanding forms of oral literature, requiring the solver to infer a concealed referent from a compressed, often metaphorical cue (Finnegan, 2012; Okpewho, 1992; Thompson, 1977). In Sepedi cultural life, dithai have historically functioned both as entertainment and as a structured training ground for observation, reasoning, and verbal agility (Maruma & Molotja, 2018). The opening call and response, Thaii!, establishes a communal, dialogic space in which a solver is expected to produce an inference, defend it, and adjust if the communal audience challenges it. Dithai therefore encode in a single short exchange, the rudiments of hypothesis generation and testing, analogical reasoning, and metacognitive regulation (Canonja, 2024; Jaxa, 2024; Makhado, 2026). Recent African studies have argued for the deliberate reintegration of riddles into formal schooling as a means of strengthening inferential and analytical capacities in learners. Wiysahnyuy and Anjirbag (2023) show that riddle-based instruction in Cameroon strengthens both comprehension and reasoning, while Idada (2025) documents similar effects in Nigerian secondary schools. Makhado (2026) argues that Tshivenḓa proverbs and riddles carry the embodied wisdom that sustains indigenous knowledge systems. In South Africa specifically, Maledu (2024) argues that Sepedi folklore, including dithai, is a significant and underused resource for teaching 21st-century skills, while Jaxa (2024) reports on the intercultural and moral affordances of folklore-based pedagogy in Further Education and Training classrooms. Despite this growing body of evidence, the specific potential of digital, gamified dithai remains largely unexplored. The Cognitive Architecture of Dithai Beyond their cultural function, dithai show a distinctive cognitive architecture that makes them especially well suited to the development of critical thinking. Every thai presents a compressed semantic puzzle in which a concealed referent is gestured at through metaphor, analogy, or structural comparison. The solver must generate candidate interpretations, test them against the literal features of the clue, eliminate incompatible options, and settle on the most economical referent that satisfies all the stated conditions. This sequence reflects the core operations of hypothesis generation, constraint satisfaction, and inference to the best explanation that feature prominently in contemporary accounts of critical thinking (Dwyer, 2017; Halpern & Dunn, 176

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

2021; Paul & Elder, 2019). The cognitive work is not displaced by memorisation, even learners who have heard a particular thai before must still reconstruct the logic of the inference to explain their answer, because the social grammar of the genre requires justification rather than mere recall. Game-based learning is now supported by a mature evidence base. Plass et al. (2015) synthesised this base and argue that well-designed educational games combine cognitive, affective, motivational, and sociocultural affordances in ways that are difficult to achieve with conventional instruction. Gee (2007) shows that games can instantiate productive regimes of competence in which learners operate at the edge of their current ability and receive immediate, informative feedback. More recent reviews consistently link well-designed educational games to improvements in critical thinking, problem-solving, and engagement (Hamari et al., 2016; Plass et al., 2015). Solehuddin et al. (2025) provide meta-analytic support for the claim that problem-based learning in which game-based learning closely resembles and enhances critical thinking in secondary-school contexts. In the South African context, Shaji (2023) argues that technology-integrated pedagogy must be designed to function under realistic infrastructure conditions and to respond to cultural specificity. Generic, culturally neutral educational games risk reproducing the very colonial epistemic patterns that South African curriculum reform has sought to disrupt (Le Grange, 2018; Ndlovu-Gatsheni, 2020; Tarisayi, 2024). The challenge, therefore, is to design games whose content, interaction patterns, and feedback structures are rooted in the indigenous cultural logic they are intended to support. Digital Storytelling, Gamification, and Indigenous Knowledge Emerging research explicitly combines educational technology with indigenous cultural content. Malliga and Balamayuranathan (2025) show that digital storytelling can help preserve and promote indigenous cultural identity in the digital age, while John and Ukpai (2025) provide a systematic review of positive effects on language learning. De Medio (2024) finds comparable critical-thinking outcomes across physical, virtual, and hybrid storytelling environments, suggesting that thoughtful digital remediation does not dilute the pedagogical value of oral traditions. Ghafar (2024) and Ghosh (2025) emphasise the emotional-cognitive integration that digital storytelling supports. These studies collectively support the core design premise of the present study, that a culturally specific game can function simultaneously as a site of cultural transmission and a site of higher-order cognitive development. Technological Pedagogical Content Knowledge Mishra and Koehler's (2006) TPACK framework provides a widely used lens for thinking about teachers' integration of technology. The framework argues that effective integration requires the simultaneous coordination of content knowledge, pedagogical knowledge, and technological knowledge. In the context of this study, the content is Sepedi dithai, the pedagogy is folklore-based critical-thinking instruction (Jaxa, 2024; Maledu, 2024), and the technology is the browser-based game deployed on Netlify. Adding a cultural dimension produces what 177

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

recent studies call culturally responsive TPACK, in which technology is chosen and configured to strengthen, rather than displace, indigenous ways of knowing (Jaiswal, 2025; Tarisayi, 2024). This framing guides the design decisions reported in the Method section. Theoretical Framework The study integrates four complementary theoretical lenses. Vygotsky’s (1978, 1987) sociocultural theory treats dithai as a cultural tool that mediates cognitive development, and dithai themselves function as semiotic resources through which more abstract cognitive operations are performed; the game mechanics function as the socially mediated scaffolding that pulls learners towards those operations. Bronfenbrenner’s (1979) ecological systems theory locates the game in the classroom microsystem while extending the mesosystem to include family members who play alongside learners at home (Rosa & Tudge, 2013). This lens draws attention to the way a well-designed cultural artefact can circulate between school and home, activating existing cultural expertise that would otherwise remain invisible to the classroom. Anderson and Krathwohl’s (2001) revised taxonomy of educational objectives provides the cognitive map used to classify dithai and gameplay tasks across the levels of remember, understand, apply, analyse, evaluate, and create. This taxonomy supplies a pragmatic alignment between cognitive ambition, task design, and assessment, and is used to classify the ten dithai in the game bank. Plass et al.’s (2015) foundations of game-based learning identify cognitive, motivational, affective, and sociocultural channels through which games influence learning, while Mishra and Koehler's (2006) TPACK framework grounds the technology integration decisions. Collectively, these frameworks support the study’s core claim: a culturally specific game can operate as a mediating cultural tool within the zone of proximal development, supported by ecological and technological affordances that teachers acting alone cannot easily provide. Method Research Design A three-iteration design-based research approach was adopted. Design-based research is wellsuited to the joint development of educational artefacts and pedagogical theory in authentic classroom contexts, and it allows iterative refinement of both the designed intervention and the claims made about it (Anderson & Shattuck, 2012; McKenney & Reeves, 2019). Iteration 1 produced the initial game prototype and was piloted with one teacher and five learners. Iteration 2 refined the riddle bank, feedback system, and scaffolding prompts on the basis of Iteration 1 observations, and was run with two teachers and ten learners. Iteration 3, the main implementation reported in this article, was conducted with four teachers and twenty learners across two primary schools in the Lebowakgomo Circuit.

178

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Participants and Setting Purposive sampling (Palinkas et al., 2015) was used to select two Sepedi-medium primary schools in the Lebowakgomo Circuit and four Sepedi home language teachers (two per school) who had taught Grade 7 for at least five years. Twenty Grade 7 learners (ten per school, stratified by self-reported device familiarity) participated in the main implementation. Ethical clearance for this study was granted by the university’s Turfloop Research Ethics Committee, and research permission was granted by the Limpopo Department of Education. The Dithai Digital Game Riddle Bank and Cognitive Mapping A bank of ten Sepedi dithai was drawn from the Grade 7 Sepedi textbook and classified using Anderson and Krathwohl’s (2001) revised taxonomy. Table 1 presents the full bank. Dithai R1 to R3 were classified at the understand and apply level because they require learners to map a clear metaphorical referent (a doorless little house, a cow in a pool, a tall weeping man) to a single concrete answer. Dithai R4 to R6 were classified at the analysis level because they require learners to decompose a structural comparison (equal blankets, two chasers, a surface that absorbs a footstep). Dithai R7 to R9 were classified at the evaluate level because they require learners to interpret a transformation or temporal phenomenon (sorghum spread at night, a cow that loses weight at the river, a container that carries its water inside it). Dithai R10 was classified at the create level because it was presented to learners as a template prompt, inviting them to compose their own short dithai modelled on the same minimal pattern. Table 1 Sepedi Dithai (Riddles) Used in the Digital Game: Answers and Cognitive Classification No.

Thai (Sepedi Riddle)

Karabo (Answer)

Cognitive Level

Ngwakwana wa koko sehlokalebati.

Lee (egg)

Understand / Apply

Kgomo ya gešo e wetše bodibeng, ka šala ke swere ka mosela.

Lefehlo (whisk)

Understand / Apply

Monna yo motelele o rothiša ditete.

Kerese (candle)

Understand / Apply

Kobo ya mme le ya tate di a lekana.

Legodimo le lefase (heaven and earth)

Analyse

Nkope le Nkope ba a lelekišana.

Maotwana a paesekele (wheels of a bicycle)

Analyse

Ka se gata sa nkgata.

Meetse (water)

Analyse

179

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

No.

Thai (Sepedi Riddle)

Karabo (Answer)

Cognitive Level

Ke anegile mabele bošego, ka re ke tsoga ka hwetša a se gona.

Dinaledi (stars)

Evaluate

Kgomo ya gešo e ile nokeng e nonne, ya boa e otile.

Sesepe (soap)

Evaluate

Nkgokolwana ya Bopedi, meetse o tšea kae?

Legapu (watermelon)

Evaluate

R10

Sese, se kae?

Letsetse (used as a template prompt for Level 4 creation)

Create

Note. Cognitive levels follow Anderson and Krathwohl (2001). Answers in English are provided in parentheses for readers unfamiliar with Sepedi. R10 was used as a template prompting learners to compose their own dithai.

Game Mechanics Learners progress through four ascending levels that correspond to the four cognitive categories in Table 1. Each level presents dithai whose cognitive demand increases systematically. Three scaffolding mechanisms are embedded. First, an adaptive hint system provides progressively more explicit clues drawn from a pre-reading prediction strategy. Second, a Reason With a Friend mechanic pairs learners for collaborative reasoning. Third, an elder-avatar feedback agent provides culturally framed explanations when a riddle is solved or missed, operationalising the mesosystem connection between home and school. All gameplay events are logged locally and synchronised to an anonymised analytics endpoint when connectivity is available. The opening call, Thaii!, is retained in the interface and is followed by the expected response, preserving the performative grammar of the genre and cueing learners into the cultural framing of the activity. Scoring follows a simple model: ten points per correct answer, a five-point streak bonus awarded once a solver records three consecutive correct answers, and a reset of the streak after an incorrect answer. Correct answers produce the acknowledgement Nnete (correct), while incorrect answers produce Aowa (no), along with the correct answer and an elder-avatar explanation. This feedback is deliberately brief, so that the cognitive work remains with the solver. Deployment through Netlify The application is version-controlled in a private Git repository and continuously deployed to Netlify. Netlify provides automatic HTTPS hosting, a global content delivery network, serverless functions for the analytics endpoint, and built-in form handling for teacher feedback. Continuous deployment meant that iterative refinements arising from Iterations 1 and 2 could be pushed to learners within minutes rather than days, materially accelerating the design-based research cycle. The game was shared with learners through a short, memorable URL and a 180

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

printable QR code that teachers distributed during class. Figures 1 to 4 illustrate the game in use: the home screen and the four cognitive levels (Figures 1–3) and the teacher analytics dashboard (Figure 4). Figure 1 Home Screen of the Dithai Digital Game (Sepedi Interface)

Note. This home screen is the learner’s entry point. It presents the ten-riddle bank, the four ascending cognitive levels (Kgato 1–4), and the Thaii! call-and-response prompt that opens each round.

Figure 2 Level 2 “Analyse” Dithai Gameplay With Adaptive Hint Open

Note. At Level 2 (Analyse), the learner selects the features that identify the concealed referent; the adaptive hint (shown open) offers a scaffolded clue without revealing the answer, extending teacher support into asynchronous play.

181

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Figure 3 Level 4 “Create” Interface With Elder-Avatar Feedback

Note. At Level 4 (Create) learners compose their own thai (call-and-response prompt) and receive feedback from an elder-avatar, modelling the communal grammar of justification that characterises the oral genre.

Figure 4 Teacher Analytics Dashboard Showing Time-On-Task and Attempts per Dithai

Note. The dashboard gives teachers per-learner time-on-task and attempts for each thai (call-and-response prompt), allowing them to identify where learners struggle and to target subsequent whole-class mediation.

182

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Data Collection Four data sources were used to support triangulation (Flick, 2018). Non-participant classroom observations, conducted by the first author, using a structured observation schedule, captured teacher mediation, peer interaction, and learner engagement. Semi-structured post-play interviews with teachers and a sub-sample of learners probed perceived cognitive demand, cultural resonance, and motivational effects. Interviews were conducted in Sepedi and English according to participant preference, audio-recorded, and transcribed verbatim. Game-play logs, automatically generated by the Netlify-hosted application, recorded time-on-task, number of attempts per dithai, use of scaffolding hints, and pathways through the game. Document analysis (Bowen, 2009) was conducted on the in-game written responses learners typed in response to the higher-order prompts at Level 3 and Level 4. Data Analysis Observation notes, interview transcripts, and learner responses were analysed using Braun and Clarke’s (2006, 2019) six-phase thematic framework, producing the four themes. Data from the game-play logs were analysed descriptively to triangulate the qualitative claims, in line with Nowell et al. (2017). Discrepant cases were examined in detail, and rival explanations were sought before themes were finalised. Trustworthiness and Ethics Lincoln and Guba’s (1985) criteria of credibility, dependability, confirmability, and transferability were addressed through triangulation across four data sources, member checking with teachers, thick description of the classroom setting, and a full audit trail (Korstjens & Moser, 2018). Informed consent was obtained from teachers and from parents or guardians, and assent was obtained from learners. All learner in-game identifiers were pseudonymised at the point of data export. Because the game collected typed responses from minors, additional data protection measures aligned with the Protection of Personal Information Act (Republic of South Africa, 2013) were implemented, including local-only storage of typed responses during play and aggregated, de-identified export for analysis. Findings Thematic analysis across the four data sources produced four principal themes. Sustained Engagement With Higher-Order Reasoning Game-log analytics showed that learners spent a mean of 27 minutes per session actively engaged with dithai, with more than 60 percent of that time at Levels 2, 3, and 4, which require analysis, evaluation, and creation. Teachers observed that learners who had previously struggled to sustain engagement with written comprehension tasks remained absorbed throughout gameplay. The compact, self-contained nature of dithai, combined with the immediate feedback loop of the game, appeared to lower the cognitive cost of sustained higher183

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

order engagement. Teacher D explained the effect, “With the game, they don't even notice that they are thinking. They just want to solve the next one. And when they solve it, they want to explain to each other why they were right.” The pattern is consistent with Gee’s (2007) argument that games create productive regimes of competence, and with Plass et al.’s (2015) claim that well-designed games integrate cognitive, motivational, and affective channels. In the present case, the cultural familiarity of the dithai appears to add a sociocultural channel that intensifies engagement further. Disaggregating the analytics by individual dithai revealed a clear cognitive-level pattern. At the understand and apply level, the dithai Ngwakwana wa koko sehlokalebati (R1) produced a median solution time of 22 seconds, while the transformation-based dithai at the evaluate level, such as Kgomo ya gešo e ile nokeng e nonne, ya boa e otile (R8), produced a median solution time of 88 seconds and attracted the largest number of hint requests. The Level 4 create prompt, in which learners were invited to compose their own dithai in the pattern of Se se, sekae (R10), produced the longest time-on-task of any single item, with a median of 142 seconds of typing and revising. These differences indicate that the game was successfully calibrated to scale cognitive demand and that learners were engaging with that scale rather than treating all dithai uniformly. Culturally Grounded Reasoning and the Re-Emergence of Botho Analysis of learners' typed in-game responses at Level 3 (evaluate) and Level 4 (create) revealed a strikingly culturally grounded moral vocabulary. Learners used concepts such as botho (humanness), kwelobohloko (compassion), and tlhompho (respect) when justifying their evaluations of dithai that involved relational referents. For example, in reasoning about the evaluate-level dithai about a cow that goes to the river fat and returns thin (R8), one learner explained that soap is like someone who practices botho, because it gives itself up in order to clean others. This metaphorical extension was not prompted by the game, but emerged spontaneously in a Level 3 free-response field. This is significant; it shows that digital remediation did not flatten the cultural depth of dithai but rather preserved and, in some cases, foregrounded it by making the evaluation component of gameplay explicit. The game provided a textual, asynchronous context that made the spontaneous emergence of botho visible and trackable. Extension of Teacher Scaffolding Into Asynchronous Modes Teachers reported that the game expanded the scaffolding they could provide. Whereas conventional classroom scaffolding is bounded by the lesson period, the Netlify-deployed game allowed learners to continue reasoning at their own pace, and enabled teachers to review game-log analytics before the next lesson. Teacher C reflected on this in the post-play interview: “I can see who is struggling at which dithai before they even tell me. So when they come to class, I already know where to help.”

184

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

This finding aligns with Vygotsky’s (1987) concept of the zone of proximal development, extending it into an asynchronous, data-informed mode of scaffolding that is difficult to achieve with oral pedagogy alone. In concrete terms, the game-log analytics allowed teachers to see, before the next lesson, exactly which dithai a given learner had attempted, how many hints they had used, and where they had stalled. This turned scaffolding from a purely in-themoment classroom act into a two-stage process: the game supported the learner during independent play, and the teacher then used the resulting data to plan the exact targeted mediation, effectively extending the learner's zone of proximal development across both home and school time. It also aligns with Mishra and Koehler’s (2006) argument that the integration of content, pedagogy, and technology produces pedagogical affordances that none of the three components can achieve on its own. Strengthening of the Home-School Mesosystem Approximately 70 percent of learners in the main implementation (n = 14 of N = 20) reported playing the game at home, often with parents, siblings, or grandparents. Several learners recounted that elders corrected their answers or contributed additional dithai that were not in the initial bank. One learner reported in a classroom observation: “My grandmother said the dithai in the game is wrong because in our village, we say it differently. So I wrote it down for the teacher.” This illustrates a live mesosystem connection in which the digital artefact becomes a boundary object (Bronfenbrenner, 1979) that circulates cultural knowledge between home and school. It also shows that digital gamification can function as a site of documentation for otherwise dispersed oral traditions, opening possibilities for learner-led curation of cultural content in future iterations of the game. Discussion From Oral Folklore to Digital Folklore: Continuity, Not Replacement A central theoretical claim of this study is that carefully designed educational technology can strengthen, rather than displace, indigenous folklore pedagogies. The four themes reported above show how the digital modality can sustain and intensify the cognitive, cultural, and pedagogical work that folklore-based instruction has long been recognised to perform (Jaxa, 2024; Maledu, 2024). The Dithai Digital Game sustained multilevel critical thinking for extended sessions, made the spontaneous use of botho in moral reasoning visible through its Level 3 and Level 4 prompts, and extended teacher scaffolding into asynchronous, datainformed modes that teachers alone could not easily achieve. A plausible explanation for these effects lies in the alignment between the cognitive structure of dithai and the mechanics of the game. Because each thai already demands inference, constraint-checking, and justification, the game did not have to impose an artificial problem structure; it simply made visible, and rewarded, reasoning that the genre itself requires. The immediate feedback loop, the ascending cognitive ladder, and the culturally familiar framing appear to have lowered the affective cost

185

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

of sustained higher-order thinking, so that learners persisted with demanding tasks that, in written form, had previously produced disengagement. Sociocultural and Ecological Reading Viewed through Vygotsky’s (1987) lens, the game functions as a mediating cultural tool in which gameplay mechanics, feedback, and peer reasoning constitute the social interaction that moves learners through their zone of proximal development. This mediating function is possible because the game externalises reasoning that would otherwise remain internal like the hint system, the elder-avatar explanation, and the Reason With a Friend mechanic each supply a more capable partner at the moment a learner reaches the limit of independent problemsolving, so the social support that Vygotsky locates in a human interlocutor is here distributed across peers, the interface, and the cultural voice of the elder. Viewed through Bronfenbrenner’s (1979) lens, the study extends the microsystem to include the digital artefact itself and strengthens the mesosystem by activating home-based cultural expertise. Plass et al.’s (2015) four channels (cognitive, motivational, affective, and sociocultural) were all present in the observed effects, but the sociocultural channel dominated, which is distinctive for an indigenous-content game. The dominance of the sociocultural channel is most plausibly explained by the cultural origins of the content itself. Unlike a generic educational game, in which motivation must be manufactured through extrinsic reward, the dithai carried preexisting cultural meaning that learners recognised from home and community life. This meant that peer reasoning, the elder-avatar feedback, and the communal call-and-response were not decorative features but activations of a social practice learners already knew, so the social and cultural dimension of play was foregrounded rather than added on. Technological Pedagogical Content Knowledge and Teacher Capacity The success of the intervention depended on teachers who combined strong content knowledge of Sepedi folklore, confident pedagogical knowledge of critical-thinking instruction, and sufficient technological knowledge to configure the game within their lessons. This is TPACK in practice (Mishra & Koehler, 2006). Variability in teachers' pedagogical repertoires is well documented (Jaxa, 2024); the present study suggests that technology integration magnifies, rather than reduces, the importance of this variability. This is most likely because the game does not replace teacher mediation but redistributes it. It extends scaffolding into asynchronous time and generates analytics that only a pedagogically skilled teacher can interpret and act upon. A teacher with a rich critical-thinking repertoire can convert the game's data into targeted whole-class mediation, whereas a teacher without that repertoire may leave those affordances unused, widening rather than narrowing the gap between classrooms. Teacher education programmes therefore need to develop culturally responsive TPACK rather than generic digital literacy (Jaiswal, 2025; Tarisayi, 2024).

186

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Implications for Educational Technology in Indigenous-Language Classrooms The study offers a replicable design pattern: take a specific, linguistically compact folklore subgenre, map its tasks onto the revised taxonomy of educational objectives, implement a mobile-first, offline-capable game, deploy continuously through a static hosting service , and embed scaffolding mechanics that mirror teachers' best face-to-face practice. This pattern could be adapted to Tshivenḓa mirero (proverbs), isiZulu izinganekwane (folktales), Xitsonga swivuriso (riddles), and related traditions across the continent. The study also offers a replicable evaluation pattern that combines game-log analytics, classroom observation, and learner responses within a design-based research cycle. These affordances carry different implications for different stakeholders. For classroom teachers, the pattern suggests that the payoff of such a game depends less on the technology than on their capacity to read its analytics and convert them into responsive mediation, which argues for professional development that pairs digital fluency with folklore-based criticalthinking pedagogy. For instructional designers, it indicates that cognitive scaffolding should be built directly onto an existing cultural task structure rather than layered over generic content, so that the game amplifies reasoning the genre already demands. For game developers, it points to the value of lightweight, offline-capable architectures and culturally specific feedback agents over high-fidelity graphics as the levers most likely to sustain higher-order engagement in resource-constrained schools. For researchers, it opens questions about how far this design logic transfers to other folklore sub-genres and languages, and about the longer-term cognitive and identity-related effects of sustained play, which the present design-based study was not positioned to measure. Limitations Several limitations should be acknowledged. Firstly, the implementation reported in this article was conducted in a single circuit, with two schools, four teachers, and twenty learners. While this scale is appropriate for a design-based research cycle that prioritises depth of insight over statistical generalisation (McKenney & Reeves, 2019), the findings should be read as analytically rather than statistically generalisable. Replication in additional Sepedi-speaking circuits, and in schools with different infrastructural and demographic profiles, is needed before broader curriculum-level recommendations can be made. Secondly, the observation window was short. The main implementation captured a single cycle of gameplay across approximately three weeks, which limits the conclusions that can be drawn about the durability of the cognitive, cultural, and pedagogical effects observed. Whether sustained engagement with higher-order reasoning persists once the novelty of the digital artefact wears off is an empirical question that this study cannot settle, and it is an obvious target for a longitudinal follow-up that tracks the same cohort over a full academic year. Thirdly, learner reports of home play depended on self-report and informal teacher follow-up rather than on instrumented logging at the home device. The figure of approximately seventy

187

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

percent of learners playing at home should therefore be read as a participant-reported estimate rather than a verified analytic count. Future iterations could incorporate optional, parentconsented home-session logging to triangulate this claim more robustly, while remaining within the data-protection commitments outlined in the Method section. Fourth, the riddle bank was deliberately small, comprising only ten dithai, to keep the cognitive ladder transparent and the implementation tractable. A larger and more dialectally diverse bank, drawn from multiple Sepedi-speaking communities, would be needed for any longerterm deployment, and would also create opportunities to investigate whether dialect variation in dithai affects the patterns of inference that learners produce. The current bank also privileges the four cognitive levels selected for the game and does not exhaust the full range of dithai used in everyday Sepedi cultural practice. Finally, the analytics endpoint captured only learner-side gameplay events. Teacher-side orchestration of the game, including when teachers paused gameplay, regrouped learners, or extended a digital prompt into oral discussion, was captured through observation rather than instrumented logging. A more integrated teacher dashboard, capturing teacher actions and decisions in vivo, would be a useful next step for both research and professional development, and would allow finer-grained analysis of how teachers exercise culturally responsive TPACK in the moment. Conclusion This study has reported on the design, deployment, and evaluation of a Sepedi dithai digital game intended to develop critical thinking in Grade 7 Home Language learners. The Netlifydeployed game sustained higher-order reasoning, preserved culturally grounded moral vocabulary, extended teacher scaffolding into asynchronous modes, and strengthened the home-school connection through multigenerational play. Theoretically, the study contributes a demonstration that sociocultural, ecological, and game-based learning frameworks can be combined to explain how digitised folklore mediates higher-order reasoning within the zone of proximal development. Practically, it offers indigenous-language teachers a concrete, low-cost design pattern that amplifies, rather than replaces, existing folklore pedagogies. Four recommendations follow. Firstly, educational technology initiatives for South African indigenous-language classrooms should begin with specific folklore sub-genres and explicit cognitive mapping, rather than with generic gamification templates. Secondly, deployment infrastructure should be chosen for realism under rural-school connectivity conditions; static hosting platforms with offline support, such as Netlify, are particularly well suited. Third, culturally responsive TPACK, rather than generic digital literacy, should become the target of teacher professional development. Fourth, future research should investigate the longitudinal cognitive, linguistic, and identity-related effects of sustained dithai-game use, and should extend the design pattern to other indigenous African languages.

188

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

In a context in which indigenous knowledge and educational technology are too often treated as competitors, this study offers empirical and design evidence that they can be made to cooperate, and that the cooperation can be both cognitively ambitious and culturally honest.

Acknowledgements The authors thank the four Sepedi Home Language teachers and twenty Grade 7 learners who participated in the design-based research cycles, the two school principals for facilitating site access, the community elders who verified the dithai bank, and the Limpopo Department of Education for granting research permission. The authors also thank colleagues in the Department of Language Education and the School of Education at the University of Limpopo for constructive feedback during the development of the game prototype. Any errors remain the responsibility of the authors. Declaration of Generative AI and AI-assisted Technologies The author(s) would like to acknowledge the use of artificial intelligence (AI) tools in the preparation of this manuscript. The following tool(s) was/were used for these specific purposes only: a large language model (OpenAI ChatGPT) was used for minor language editing and proofreading, and to assist with drafting portions of the HyperText Markup Language, Cascading Style Sheets, and JavaScript code used to build the browser-based Dithai Digital Game. All AI-edited text was thoroughly reviewed and revised by the authors to ensure accuracy, clarity, and adherence. AI tools were not used for other purposes, including data generation, data analysis, methodology, or interpretation of findings or conclusions. The author(s) take(s) full responsibility for all aspects of the final manuscript.

189

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

References Alsaleh, N. J. (2020). Teaching critical thinking skills: Literature review. Turkish Online Journal of Educational Technology, 19(1), 21–39. Anderson, L. W., & Krathwohl, D. R. (2001). A taxonomy for learning, teaching, and assessing: A revision of Bloom's taxonomy of educational objectives. Longman. Anderson, T., & Shattuck, J. (2012). Design-based research: A decade of progress in education research? Educational Researcher, 41(1), 16–25. https://doi.org/10.3102/0013189X11428813 Bowen, G. A. (2009). Document analysis as a qualitative research method. Qualitative Research Journal, 9(2), 27–40. https://doi.org/10.3316/QRJ0902027 Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101. https://doi.org/10.1191/1478088706qp063oa Braun, V., Clarke, V., & Hayfield, N. (2019). A starting point for your journey, not a map: Nikki Hayfield in conversation with Virginia Braun and Victoria Clarke about thematic analysis. Qualitative Research in Psychology, 19(2), 424–445. https://doi.org/10.1080/14780887.2019.1670765 Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and design. Harvard University Press. Canonja, M. L. (2024). Reviving African indigenous education via folktales: Pedagogical integration, cultural relevance, and philosophical underpinnings in modern educational systems. E-Journal of Humanities, Arts and Social Sciences, 5(12), 1–15. https://doi.org/10.38159/ehass.20245121 De Medio, C. (2024). Digital storytelling and critical thinking: A comparative analysis across physical, virtual, and hybrid learning environments. Ubiquity Proceedings, 4(1), 41– 41. https://doi.org/10.5334/uproc.163 Dwyer, C. P. (2017). Critical thinking: Conceptual perspectives and practical guidelines. Cambridge University Press. https://doi.org/10.1017/9781316537411 Finnegan, R. (2012). Oral literature in Africa. Open Book Publishers. https://doi.org/10.11647/OBP.0025 Flick, U. (2018). An introduction to qualitative research (6th ed.). Sage. Gee, J. P. (2007). What video games have to teach us about learning and literacy (2nd ed.). Palgrave Macmillan. Ghafar, Z. (2024). Storytelling as an educational tool to improve language acquisition: A review of the literature. Journal of Digital Learning and Distance Education, 2(10), 781–790. https://doi.org/10.56778/jdlde.v2i9.227 Ghosh, P. (2025). The role of storytelling in enhancing emotional and cognitive development in children. International Journal for Multidisciplinary Research, 7(5), 1–3. https://doi.org/10.36948/ijfmr.2025.v07i05.58403 Halpern, D. F., & Dunn, D. S. (2021). Critical thinking: A model of intelligence for solving real-world problems. Journal of Intelligence, 9(2), 22. https://doi.org/10.3390/jintelligence9020022

190

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Hamari, J., Shernoff, D. J., Rowe, E., Coller, B., Asbell-Clarke, J., & Edwards, T. (2016). Challenging games help students learn: An empirical study on engagement, flow and immersion in game-based learning. Computers in Human Behavior, 54, 170–179. https://doi.org/10.1016/j.chb.2015.07.045 Howie, S. J., Combrinck, C., Roux, K., Tshele, M., Mokoena, G. M., & McLeod Palane, N. (2017). PIRLS Literacy 2016: South African highlights report. Centre for Evaluation and Assessment. Idada, I. (2025). Exploring the impact of indigenous storytelling on critical thinking skills development among secondary school students in Edo State, Nigeria. Advances in Social Science, Education and Humanities Research, 151–158. https://doi.org/10.2991/978-2-38476-527-0_10 Jaiswal, A. (2025). Indigenous pedagogies: Teaching and learning practices rooted in local contexts. Naveen International Journal of Multidisciplinary Sciences, 1(4), 97–104. https://doi.org/10.71126/nijms.v1i4.40 Jaxa, N. P. (2024). Enhancing indigenous knowledge systems in education through folklore: A case study at the Further Education and Training Phase in South Africa. Journal of Culture and Values in Education, 7(4), 132–148. https://doi.org/10.46303/jcve.2024.45 John, E. A., & Ukpai, G. (2025). Digital storytelling: A systematic review of its impact in language education. Asian Journal of Applied Education, 4(4), 461–472. http://dx.doi.org/10.55927/ajae.v4i4.14698 Korstjens, I., & Moser, A. (2018). Series: Practical guidance to qualitative research. Part 4: Trustworthiness and publishing. European Journal of General Practice, 24(1), 120– 124. https://doi.org/10.1080/13814788.2017.1375092 Le Grange, L. (2018). Decolonising, Africanising, indigenising, and internationalising curriculum studies. Journal of Education, 74, 4–18. https://doi.org/10.17159/2520-9868/i74a01 Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. Sage. Makhado, A. J. (2026). Embodied wisdom and cultural memory: The role of Tshivenḓa proverbs in sustaining indigenous knowledge systems. Southern African Journal for Folklore Studies, 35(1), 1–19. https://doi.org/10.25159/2663-6697/20047 Maledu, A. D. (2024). The significance of Sepedi folktales in teaching 21st century skills. Journal for Language Teaching, 58(1). https://doi.org/10.56285/jltvol58iss1a6538 Malliga, N., & Balamayuranathan, B. (2025). Digital storytelling as a medium for preserving and promoting indigenous cultural identity in digital age. Shanlax International Journal of Arts Science and Humanities, 12(S3-Apr), 63–69. https://doi.org/10.34293/sijash.v12iS3-Apr.9054 Maruma, M. W., & Molotja, T. W. (2018). The relevance of folklore in an indigenous language teaching and learning situation: The case study of Sepedi. Southern African Journal for Folklore Studies, 28(1), 11 pages. https://doi.org/10.25159/1016-8427/4293 McKenney, S., & Reeves, T. C. (2019). Conducting educational design research (2nd ed.). Routledge. https://doi.org/10.4324/9781315105642

191

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Mishra, P., & Koehler, M. J. (2006). Technological pedagogical content knowledge: A framework for teacher knowledge. Teachers College Record, 108(6), 1017–1054. https://doi.org/10.1111/j.1467-9620.2006.00684.x Ndlovu-Gatsheni, S. J. (2020). Decolonisation as repedagogisation: Untying the knots of coloniality from education. Postcolonial Directions in Education, 9(2), 171–194. https://www.um.edu.mt/library/oar/handle/123456789/65670 Nowell, L. S., Norris, J. M., White, D. E., & Moules, N. J. (2017). Thematic analysis: Striving to meet the trustworthiness criteria. International Journal of Qualitative Methods, 16(1), 1–13. https://doi.org/10.1177/1609406917733847 Okpewho, I. (1992). African oral literature: Backgrounds, character, and continuity. Indiana University Press. Palinkas, L. A., Horwitz, S. M., Green, C. A., Wisdom, J. P., Duan, N., & Hoagwood, K. (2015). Purposeful sampling for qualitative data collection and analysis in mixed method implementation research. Administration and Policy in Mental Health, 42(5), 533–544. https://doi.org/10.1007/s10488-013-0528-y Paul, R., & Elder, L. (2019). The thinker's guide to analytic thinking: How to take thinking apart and what to look for when you do (3rd ed.). Rowman & Littlefield. Plass, J. L., Homer, B. D., & Kinzer, C. K. (2015). Foundations of game-based learning. Educational Psychologist, 50(4), 258–283. https://doi.org/10.1080/00461520.2015.1122533 Pretorius, E. J., & Spaull, N. (2022). Reading research in South Africa (2010–2022): Coming of age and accounting for empirical regularities. In N. Spaull & E. J. Pretorius (Eds.), Early grade reading in South Africa (pp. 11–32). Oxford University Press. Republic of South Africa. (2013). Protection of Personal Information Act (No. 4 of 2013). Government Printer. Rosa, E. M., & Tudge, J. (2013). Urie Bronfenbrenner's theory of human development: its evolution from ecology to bioecology. Journal of Family Theory & Review, 5(4), 243–258. https://doi.org/10.1111/jftr.12022 Shaji, G. (2023). Preparing students for an AI-driven world: Rethinking curriculum and pedagogy in the age of artificial intelligence. Partners Universal Innovative Research Publication, 1(2), 112–136. https://doi.org/10.5281/zenodo.10245675 Solehuddin, M., Sartinayanti, S., Hasnah, S., Karmila, M., & Muriyanto, M. (2025). Enhancing critical thinking skills through problem-based learning: A classroom intervention in junior high school. International Journal of Educational Research Excellence, 4(1), 338–344. https://doi.org/10.55299/ijere.v4i1.1390 Tarisayi, K. S. (2024). Integrating indigenous knowledge in South African geography education curricula for social justice and decolonization. E-Journal of Humanities, Arts and Social Sciences, 5(7), 1195–1206. https://doi.org/10.38159/ehass.20245711 Thompson, S. (1977). The Folktale. University of California Press. Thornhill-Miller, B., Camarda, A., Mercier, M., Burkhardt, J. M., Morisseau, T., BourgeoisBougrine, S., Vinchon, F., El Hayek, S., Augereau-Landais, M., Mourey, F., Feybesse, C., Sundquist, D., & Lubart, T. (2023). Creativity, critical thinking, communication, and collaboration: Assessment, certification, and promotion of 21stcentury skills for the future of work and education. Journal of Intelligence, 11(3), Article 54. https://doi.org/10.3390/jintelligence11030054 192

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Trilling, B., & Fadel, C. (2009). 21st century skills: Learning for life in our times. JosseyBass. Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Harvard University Press. Vygotsky, L. S. (1987). Thinking and speech. Plenum Press. Wiysahnyuy, L. F., & Anjirbag, M. A. (2023). Folktales as indigenous pedagogic tools for educating school children: A mixed methods study among the Nso of Cameroon. Frontiers in Psychology, 14(1049691). https://doi.org/10.3389/fpsyg.2023.1049691 Corresponding author: Nkame Emmanuel Ngobeni Email: nkamengobeni@gmail.com

193

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

194

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Four-Domain Framework for Evaluating Education Management Information System Effectiveness in Bhutan Gembo Tshering Paro College of Education, Royal University of Bhutan, Bhutan Saroj Thapa Royal Academy, Druk Gyalpo’s Institute, Bhutan Tenzin Choden Lekphell Paro College of Education, Royal University of Bhutan, Bhutan Chimi Dema Paro College of Education, Royal University of Bhutan, Bhutan

195

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Abstract Bhutan’s education system has made significant strides in terms of access and infrastructure development. Nevertheless, enhancing the quality of learning continues to pose a persistent challenge, despite the implementation of the Education Management Information System (EMIS) to monitor factors that impact learning quality. This study evaluates the effectiveness of Bhutan’s Education Management Information System (EMIS) and the Motherboard platform using the Systems Approach for Better Education Results (SABER-EMIS) framework. Drawing on a cross-sectional evaluative design and a cross-sectional survey of 281 schools, 20 district education officers, and national-level stakeholders, the study identifies discrepancies between intended system design and operational performance. The findings are synthesized into a four-domain evaluation framework encompassing (a) enabling environment, (b) system soundness, (c) data quality, and (d) data utilization in decision-making. Results indicate that while data quality and utilization are approaching established levels, enabling conditions and system integration remain at emerging stages, constraining the effectiveness of data-driven educational management. The study contributes a transferable evaluation model for assessing EMIS effectiveness and highlights the value of integrating quantitative administrative data with qualitative learning analytics. Based on the findings, the paper proposes a phased roadmap combining short-term improvements in data validation and accessibility with long-term strategies for interoperability, capacity development, and international benchmarking. The framework and findings offer practical and scalable insights for strengthening education data systems in low- and middle-income contexts. Keywords: Bhutan, data quality, data-driven decision-making, education information systems EMIS, Motherboard, SABER-EMIS

196

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Bhutan’s education system has made notable progress since the 1950s, achieving a net enrolment rate of 93.3% for primary education, with one school per approximately 269 households (National Statistics Bureau, 2018; Policy and Planning Division, 2022; Tshering & Bahl, 2023; Tshering & Choden, 2023). In comparison, the United States reports one school per 962 households (Kositsky, 2023; United States Census, 2021). This higher density of schools in Bhutan may reflect efforts to improve accessibility in rural areas, while the lower graduation rate highlights ongoing challenges in student retention compared to the United States. Despite these advancements, Bhutan faces persistent challenges in ensuring equitable and high-quality education. For instance, its high school graduation rate is 82.0% (Bhutan Council for School Examinations and Assessment, 2021), lower than the United States’ 88.9% (United States Census, 2021). Key challenges include disparities between urban and rural regions and limited resources for teacher training. The Education Management Information System (EMIS) helps address these by identifying gaps and tracking progress over time, supporting targeted interventions. Addressing these challenges requires leveraging education data to inform decisions, improve outcomes, and establish robust accountability mechanisms, as seen in global practices (Global Partnership for Education, 2019; National Forum on Educational Statistics, 2021; UNESCO, 2019). To strengthen evidence-based decision-making and improve student learning, Bhutan relies on the EMIS, designed to collect, manage, and analyse education data for monitoring and policy development (Abdul-Hamid, 2017; van Wyk & Crouch, 2020; World Bank Group, 2011). Recent trends in Bhutan’s EMIS implementation, such as its role in supporting resource allocation and monitoring regional disparities, further demonstrate its growing impact on education policy and practice. Bhutan’s journey to establishing its EMIS, a platform designed to collect, manage, and analyse educational data, began in the early 2000s and led to its first operational launch in 2010 with international support (Ministry of Education and Skills Development, n.d.). EMIS enables schools, districts, and policymakers to access accurate information about enrolment, resources, and performance, supporting evidence-based decisions to improve educational outcomes (Abdul-Hamid, 2017; Global Partnership for Education, 2019; UNICEF, 2019; van Wyk & Crouch, 2020; World Bank Group, 2011). Alongside EMIS, Bhutan has developed a homegrown platform called the Motherboard (Druk Gyalpo’s Institute, n.d.). The Motherboard is a locally developed system focused on qualitative assessment within the Bhutan Baccalaureate model, which emphasises holistic education and criterion-based descriptive reporting. The Bhutan Baccalaureate is a schooling approach that values a broader set of learning outcomes beyond traditional exams, including formative assessment and comprehensive feedback on students’ progress. Currently, the Motherboard is implemented in selected schools with plans for national expansion (Office of the Prime Minister and Cabinet, 2021). While EMIS primarily tracks numerical data, the Motherboard captures qualitative learning evidence, such as changes in students’ learning journeys across different cycles. Both EMIS and the Motherboard rely on ICT. Information and Communication Technology (ICT) refers to tools and resources—like computers, the internet, and smartphones—that facilitate communication and information processing in education. Bhutan has taken significant steps to integrate ICT into its education system through policies

197

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

and initiatives such as “Towards Digitised Bhutan” (National Statistics Bureau, 2020), “Leverage ICT for Learning” (Ministry of Education, 2014), and the “iSherig” plans (Ministry of Education, 2019). These efforts have resulted in improved internet connectivity, greater computer access in schools, and widespread smartphone ownership (National Statistics Bureau of Bhutan, 2018; Policy and Planning Division, 2022; Tshering & Phuntsho, 2025). A notable outcome is the Bhutan Professional Standards for Teachers, which now require teachers to use ICT as part of their teaching practice, leading to measurable increases in digital literacy and EMIS adoption across the country. Despite their complementary strengths, EMIS and Motherboard remain neither integrated nor systematically evaluated for their roles in supporting data-based decision-making and interoperability within Bhutan’s education system. This fragmentation limits the ability to leverage both quantitative and qualitative data in a coherent manner, thereby constraining informed and context-responsive educational planning. At the same time, EMIS implementation continues to face persistent operational challenges that undermine its reliability and effectiveness, including inconsistent data entry across schools, limited technical support, and recurrent system breakdowns (Ministry of Education and Skills Development, n.d.; Poudel, 2022). Consequently, data are often incomplete or inaccurate, reducing their utility for identifying educational gaps, guiding resource allocation, and monitoring progress. These limitations are particularly consequential in contexts where timely and reliable evidence is required for policy development and targeted interventions (van Wyk & Crouch, 2020). In light of these limitations, there is a clear need for an approach that both strengthens data quality and enhances system integration. The MIEMIS project, Enhancing Data Utilization and Decision-Making in Bhutan’s Education System: Motherboard Integrated Education Monitoring and Information System (MIEMIS), of which this study is a part, addresses this need by investigating the potential for integrating EMIS and Motherboard and assessing their combined impact on enhanced data utilization and decision-making. Contextual Background As this study is a component of the broader MIEMIS project, the following section provides contextual background by outlining the project’s objectives, research questions, and methodology. This section aims to position this study within the larger context of the MIEMIS project. MIEMIS Objectives The overarching objective of MIEMIS is to advance data systems and data utilization in Bhutan’s education system. This objective will be achieved through the following specific objectives: (a) Generate evidence-based insights on scaling EMIS and the Motherboard.

198

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

(b) Enhance stakeholder capacity building for knowledge utilization and EMIS and Motherboard. (c) Mobilize knowledge and evidence generated on EMIS and Motherboard for informed policy and practice. (d) Enhance stakeholder engagement and foster partnerships and collaboration for sustainable impact. (e) Strengthen the capacity of EMIS to address Gender Equality and Inclusion issues effectively. MIEMIS Research Questions (a) How can EMIS and Motherboard be effectively adapted, scaled, and implemented to promote gender equality, equity, and inclusion in education? (b) What are the enabling factors, barriers, and incentives influencing the scaling of EMIS V3.0 and the Motherboard in data systems and data used for education? (c) How can EMIS, Motherboard, artificial intelligence, and social-media platforms effectively enhance data utilization and decision-making processes in education? (d) How can data use be expanded and diversified to promote public accountability while improving educational outcomes? (e) How can EMIS and Motherboard be interoperable and tested to ensure their suitability and impact in the Bhutanese school education system? (f) How can EMIS be enhanced to incorporate gender-responsive features and practices, ensuring inclusivity and addressing gender-specific needs within the education sector? MIEMIS Method The MIEMIS project adopts action research–driven design (McNiff & Whitehead, 2006; Mills, 2014; Stringer et al., 2010), structured across six interconnected phases. This methodology is illustrated in Figure 1, which aligns the stages of action research with project phases and work packages (WP5–WP11; WP stands for Work Package).

199

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Figure 1 Action Research Elements vis-vis the Project Phases, and Work Packages

The cycle begins with Observe (Phase 1; WP5), where a situational analysis establishes baseline evidence, assesses EMIS and Motherboard system soundness, and examines integration feasibility using the SABER-EMIS framework. This is followed by Reflect (Phase 2; WP6), where stakeholders review findings and co-design gender-responsive interventions and capacity-building strategies. The present study focuses exclusively on the Observe phase, which provides the baseline evidence for subsequent phases of the MIEMIS project. The Act stage (Phase 3; WP7) implements the selected interventions, while Evaluate (Phase 4; WP8) assesses their effectiveness using established evaluation and gender-responsive frameworks (International Development Research Center, n.d.; Rogers, 2014; United Nations Girls’ Education Initiative, 2010; UNICEF, 2019), with a focus on data quality, system usability, and evidence-informed decision-making. The findings inform Modify (Phase 5; WP9), where interventions are refined using adaptive learning and scaling approaches (Begimkulov & Darr, 2023; McLean & Gargani, 2019). Finally, move in new directions (Phases 6; WP10–WP11) focuses on scaling, sustainability, partnerships, and dissemination. Continuous feedback and stakeholder engagement ensure iterative improvement throughout the project. Comprehensive Situational Analysis of EMIS and Motherboard As shown in Figure 1, Phase 1 (WP5) of the MIEMIS project, the Observe phase of action research, focuses on conducting a situational analysis of Bhutan’s existing Education Management Information System (EMIS) and the Motherboard. The situational analysis is led by the MIEMIS project team in collaboration with Bhutan’s Ministry of Education and Skills Development and other relevant stakeholders, ensuring broad engagement and impact across the education sector. This phase aims to provide a comprehensive assessment of EMIS and Motherboard components, collect and analyse baseline data, and identify data quality and

200

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

integrity issues in the two systems. Specifically, this study aims to (a) assess the current use and limitations of EMIS and Motherboard, (b) explore the feasibility and mechanisms of their integration, and (c) identify the conditions and mechanisms needed for integration to enhance data-informed decision-making in Bhutan’s education system. Literature Review The effectiveness of Education Management Information Systems (EMIS) has been widely examined as a critical factor in strengthening education governance, improving accountability, and enabling data-driven decision-making. Globally, EMIS are positioned not merely as data repositories but as integrated systems that transform raw data into actionable insights for policy and practice. The literature consistently emphasises that the value of EMIS lies in its ability to connect data production, system functionality, and decision-making processes within a coherent ecosystem (UNESCO, 2019; World Bank Group, 2011). EMIS as a Systems-Based Approach The EMIS approach aligns with the Systems Approach for Better Education Results (SABEREMIS), a comprehensive framework developed by the World Bank Group to evaluate the effectiveness of EMIS across four policy domains: Enabling environment, system soundness, data quality, and utilization (Abdul-Hamid, 2017; Abdul-Hamid et al., 2017). Research further indicates that weak linkages between these domains often result in fragmented systems, where data are collected but not effectively used (van Wyk & Crouch, 2020). For instance, countries with relatively strong technical systems but limited institutional capacity tend to struggle with translating data into policy action. This reinforces the importance of viewing EMIS as a holistic system rather than a purely technical infrastructure. The enabling environment—comprising legal frameworks, governance structures, infrastructure, and human resources—is consistently identified as a foundational determinant of EMIS performance. Strong policy frameworks and sustained financial investment are necessary to institutionalise data practices and ensure system sustainability (Global Partnership for Education, 2019). However, the literature highlights that many low- and middle-income countries face persistent challenges in this domain, including limited technical expertise, inadequate professional development, and weak data cultures (Abdul-Hamid et al., 2017; Iyengar et al., 2026). These constraints often lead to underutilisation of EMIS, even when systems are technically functional (UNICEF, 2019). Capacity building, particularly in data literacy and leadership, is therefore emphasised as a critical intervention for improving EMIS outcomes. System soundness refers to the technical robustness and architectural coherence of EMIS, including data integration, coverage, and adaptability. The literature underscores interoperability—the ability of EMIS to interact with other data systems—as a key dimension of system effectiveness (UNESCO, 2019; UNESCO, 2022). Emerging research highlights the growing need for integrated data ecosystems that combine administrative, financial, and

201

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

learning data. Countries that have successfully implemented interoperable systems demonstrate improved efficiency, reduced duplication, and enhanced policy responsiveness (Mollerus et al., 2025; UNESCO, 2022). Conversely, fragmented architectures often result in inconsistent data flows and increased reporting burdens on schools. Data quality, encompassing accuracy, reliability, timeliness, and integrity, is a central concern in EMIS research. High-quality data are essential for building trust among users and ensuring that decisions are based on credible evidence. Frameworks such as the Education Data Quality Assessment Framework (Ed-DQAF) emphasise standardisation, validation mechanisms, and clear methodological guidelines (Abdul-Hamid, 2017). Despite these frameworks, empirical studies frequently report issues such as incomplete data, inconsistencies across reporting levels, and delays in data submission (Adhikari & Budhathoki, 2025; Ariko, 2024, International Development Research Center, 2024). These challenges are often linked to human factors, including limited training and high workloads at the school level. As a result, improving data quality requires both technical solutions and organisational change. Effective Utilisation: While many countries have made progress in data collection, the utilisation of EMIS data remains uneven. The literature identifies a persistent “data-use gap,” where data are generated but not systematically used for decision-making (World Bank Group, 2011). Effective utilisation depends on accessibility, user capacity, and the relevance of data outputs to policy and practice. Studies show that when EMIS outputs are presented in userfriendly formats and integrated into routine workflows, they are more likely to inform planning, monitoring, and evaluation (UNESCO, 2018; van Wyk & Crouch, 2020). Conversely, complex interfaces and limited dissemination strategies hinder uptake. Integrating Quantitative and Qualitative Data Systems Recent advancements in education data systems highlight the importance of integrating quantitative EMIS data with qualitative learning analytics (Abu et al., 2026). Traditional EMIS primarily capture administrative and numerical data, such as enrolment, attendance, staffing, and examination results, which are valuable for monitoring system performance but provide limited insights into teaching and learning processes. Emerging platforms, such as the Motherboard, completement these datasets by capturing richer qualitative information, including formative assessment, student feedback, classroom practices, and learner engagement. Together, these data sources provide a more comprehensive understanding of educational quality and learner outcomes. The integration of quantitative and qualitative data offers considerable potential to strengthen evidence-informed decision-making at classroom, school, and system levels (Abu et al., 2026). By combining administrative records with learning and instructional data, educators and policymakers can identify learning gaps, monitor student progress, evaluate interventions, and design more targeted improvement strategies. Such integrated systems also support continuous monitoring, personalised learning, and more responsive educational planning.

202

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Despite these benefits, the literature identifies several implementation challenges, including inconsistent data standards, interoperability issues between platforms, concerns regarding data quality and governance, and the increased technical and organisational complexity associated with managing integrated systems (International Development Research Center, 2024; Mbawala et al., 2024). Addressing these challenges requires robust governance frameworks, technical capacity, stakeholder collaboration, and sustained investment to ensure that integrated education data systems effectively support teaching, learning, and policy development. Synthesis and Research Gap The literature converges on the understanding that EMIS effectiveness is shaped by the interaction of enabling conditions, system design, data quality, and data use. While existing frameworks such as SABER-EMIS provide robust tools for assessment, there remains a gap in studies that examine the integration of multiple data systems—particularly those combining quantitative and qualitative data—in low- and middle-income contexts. In the case of Bhutan, limited research has explored how EMIS and complementary platforms such as the Motherboard can be evaluated within a unified framework. This study addresses this gap by using SABER-EMIS four-domain evaluation model not only to assess the performance of the existing EMIS but also to evaluate the Motherboard’s potential for integration with EMIS, thereby enhancing a more comprehensive and evidence-informed approach to data-driven decision-making. Methodology Although the broader MIEMIS project is guided by an action research methodology, this paper reports only the baseline situational analysis undertaken during the Observe phase. Accordingly, the study employed a cross-sectional evaluative design (Maier et al., 2023) to assess the status of Bhutan’s Education Management Information System (EMIS) using the SABER-EMIS framework. Conceptual Framework for Situational Analysis Study This study employs the Systems Approach for Better Education Results–Education Management Information Systems (SABER-EMIS) framework shown in Figure 2 (AbdulHamid, 2014) to guide the assessment of Bhutan’s EMIS within the school education context. The SABER-EMIS framework structures its analysis around four interrelated policy areas, each critical to the overall effectiveness of an education information system: (a) Enabling Environment, which evaluates the presence of supportive policies, sustainable infrastructure, and human resources for robust data collection and management; (b) System Soundness, focusing on how well system processes and structures underpin comprehensive information management; (c) Quality Data, which assesses the accuracy, reliability, and timeliness of information generated by the EMIS; and (d) Utilisation for Decision-Making, which examines the practical application of EMIS data in policy and operational processes. The following table

203

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

presents the specific indicators associated with each policy area, which form the basis for systematic evaluation and targeted recommendations. Figure 2 SABER-EMIS Approach* Policy Area

Set of Indicators 1. Legal framework SWOT 2. Organizational structure and institutionalized processes Enabling 3. Human resources Environment 4. Infrastructural capacity 5. Budget Strengths Weaknesses 6. Data-driven culture 7. Data architecture 8. Data coverage System 9. Data analytics soundness 10. Dynamic system Opportunities Threats 11. Serviceability 12. Methodological soundness 13. Accuracy and reliability Quality data 14. Integrity 15. Periodicity and timeliness 16. Openness to EMIS users Utilization for 17. Operational use 18. Accessibility decision19. Effectiveness in disseminating making findings/results *For more detailed information about the policy areas and their associated factors, refer to Abdul-Hamid (2014).

The SABER-EMIS framework’s structure reflects a logical progression and interdependency between policy areas. A strong enabling environment—characterised by robust legal frameworks, sufficient resources, and a data-driven culture—lays the groundwork for system soundness by ensuring the necessary supports for effective data architecture, coverage, and analytics. In turn, a sound system is essential for generating high-quality data, as the reliability and integrity of information depend on both structural and procedural rigour. Ultimately, the presence of quality data underpins the meaningful utilisation of EMIS outputs for decisionmaking, ensuring that educational policies and practices are informed by accurate, timely, and accessible information. Thus, improvements in enabling environment indicators (such as budget and infrastructure) are likely to cascade through system soundness, data quality, and ultimately enhance data-driven decision-making capacity. Each policy area is systematically assessed through its respective indicators, with findings subjected to SWOT (Strengths, Weaknesses, Opportunities, Threats) analysis (Humphrey, 2005). This integrated process allows assessors to identify strengths and areas for improvement across the EMIS landscape, forming a basis for targeted action plans. The SABER-EMIS approach draws on established standards such as ISO 9000, the Education Data Quality

204

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Assessment Framework (Ed-DQAF), and Utilisation-Focused Evaluation (UFE), supporting its validity and reliability (Abdul-Hamid, 2014). Method of Data Collection Stakeholder Mapping and Participant Selection Stakeholders were identified and mapped using the Power-Interest Matrix (Bryson, 2004; Maqbool et al., 2022). The MIEMIS project team collaboratively identified EMIS stakeholders and allocated them to the relevant quadrants based on their decision-making authority over EMIS policies (power) and their level of engagement with EMIS data and processes (interest), as determined through team discussions and review of organisational roles (Tshering, 2025a; 2025b; 2025c; 2025d). This approach helped ensure transparency and clarity in the allocation process. The resulting stakeholder mapping is illustrated in Figure 3. As shown in Figure 3, stakeholders were grouped according to their power/influence and interest relationship, whether it involved data use, data sourcing, management, or benefiting from data. The top-right quadrant identified stakeholders and organisations with both strong influence and high interest in EMIS and Motherboard. GovTech, MoESD, ECCD, PCE, DGI, NGOs, IDRC, Dzongkhag Administration, and ROM Tech were among those classified as having high influence and interest, based on their direct engagement with EMIS and Motherboard. For example, GovTech oversees technical infrastructure, while MoESD is responsible for policy direction. NGOs and IDRC contribute through capacity building and research support. ECCD and PCE are involved in early childhood and curriculum development, DGI manages the Motherboard, Dzongkhag Administration supports implementation at the district level, and ROM Tech provides technical solutions. This classification provides additional context for readers by clarifying each entity’s specific role or contribution to the EMIS landscape.

205

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

High

Figure 3 Power-Interest Matrix ECCD Proprietors/Center Directors

Government & Regulatory Bodies

MOESD

Schools Principals & Admins

PCE

Satisfy

Manage

NGOs

DGI GovTech & ROMTech Power & Influence

Policy Makers

IDRC

Dzongkhag Administrations

ECCD Facilities RUB

Monastic Education

Teachers Local Administrations

Researchers & Academics

Monitor

Inform International Organizations

MoH Low

Students MoHCA Low Interest

Parents & Guardians

CSOs High

Sampling Frame and Sample Size Determination Data for this study were collected in March 2023. At that time, there were 45 ECCD centers, 331 primary schools, 42 lower secondary schools, 57 middle secondary schools, and 93 higher secondary schools, bringing the total to 568 schools (Policy and Planning Division, 2022). After outlining the total number of schools within our sampling frame, we proceeded to determine an appropriate sample size using the following formula. !.# ! .$(&'$)

𝑛 = (!'&)) ! *# ! .$(&'$)

Equation (1)

Where: • N=568 • Z=1.96 (for 95% confidence level) • p=0.5 (maximum variability) • E=0.05 (margin of error) The sample size was determined to be 229 schools based on 95% confidence level, maximum variability (p=.05), and 0.05 margin of error. To ensure adequate representation and account for potential non-responding schools, we increased this sample size by 52 schools, representing an anticipated non-response rate of approximately 23%. This adjustment resulted in a final sample size of 281 schools.

206

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Participant Selection and Response Rate The sample sizes for various schools were determined by multiplying the ratio of their respective number to the total number of schools by the 281 schools. Table 1 presents the sample sizes for each school. Table 1 Sample Sizes of Different Schools School ECCD PS LSS MSS HSS Total

Number (n) 525 333 39 59 93

Ratio (281/n) 0.50 0.32 0.04 0.06 0.09

Sample Size 140 89 10 16 25

Adjusted Ratio 0.10 0.57 0.07 0.10 0.16

Adjusted Sample Size 28 161 19 29 45 282

The initial sample size was adjusted using a weighted approach, with weights determined by consensus among the MIEMIS team. Weights were assigned based on criteria including school size, heterogeneity, and homogeneity; for example, larger schools received higher weights to ensure proportional representation. The adjusted sample size serves as the final sample size for the study. Bhutan comprises 20 Dzongkhags (districts), each with distinct types of schools. Consequently, the final adjusted sample sizes were used to select the number of schools from each Dzongkhag. With the adjusted sample in place, within-school sample types were selected using simple random sampling (a method where each member has an equal chance of being selected). A total of 281 schools across 20 Dzongkhags and three Thromdes (urban municipalities) were sampled for the study. In addition to schools, stakeholders also participated in the study. The stakeholders included Dzongkhag and Thromde Education Officers, Government Technology EMIS Officers, PPD, and EMO Officials. Their participation was contingent upon their experience with the EMIS. The response rate, which is calculated as the number of completed survey questionnaires received divided by the total number of individuals invited to participate, was 75% for schools and 62% for Dzongkhag and Thromde Education Officers. The response rates for PPD, EMO Officials, Government Technology, EMIS Officers, and the Department of School Education were all 100%. For the situational analysis study of the Motherboard, data were collected from the 24 Bhutan Baccalaureate schools where the Motherboard is utilized. A survey was conducted involving 24 principals, 30% of teachers from each school and 6% of students from grades 5 to 12. The survey also included system administrators from Druk Gyalpo’s Institute.

207

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Data Collection Tool The study utilised the SABER-EMIS questionnaire to gather data. SABER-EMIS (Systems Approach for Better Education Results-Education Management Information Systems) is a framework developed by the World Bank to assess the effectiveness of education data systems. The SABER-EMIS questionnaire was selected due to its comprehensive approach to evaluating education management information systems, aligning closely with the study’s aim to assess system implementation at multiple administrative levels. The questionnaire (Abdul-Hamid, 2014) assessed the 19 indicators enumerated in Table 3. Each indicator is further subdivided into sub-indicators, each comprising a set of individual items. Each item is presented as a dichotomous question, necessitating a “Yes” or “No” response. Notably, if the respondent answers “Yes” to an item, they are prompted to specify the level at which that particular aspect is implemented. This study delineates the pertinent levels as Ministry, Dzongkhag, and School. ‘Dzongkhag’ refers to an administrative district in Bhutan. To elucidate the format of these items, consider the following illustration: “Is there a law to establish or create an education management information system (EMIS) that collects, processes, and disseminates education data on a regular basis?” This example exemplifies the structure of the items within the SABER-EMIS questionnaire, emphasising the dichotomous response options and the subsequent level selection for “Yes” responses. Table 3 Policy Areas, Indicators, and Sub-Indicators in the SABER-EMIS Questionnaire Policy Area

Enabling Environment

Indicator

Legal Framework

Sub-Indicator (Number of Items) Institutionalization (4) Responsibility (2) Dynamic framework (2) Data Supply (7) Comprehensive and quality care (7) Data sharing and coordination (2) Utilization (1) Budget (2) Confidentiality (15)

Organizational Structure and Institutional Processes

(6)

Human Resources

Infrastructure Capacity

Budget

Personnel (3) Professional development (11) Data collection means (3) Database (2) Data management system (11) Data dissemination means (2) Personnel and professional development (4) Maintenance (1) Reporting (1)

208

IAFOR Journal of Education: Technology in Education

Data-driven culture Data architecture Data coverage

System Soundness

Data analytics Dynamic system

Serviceability

Methodological soundness

Quality Data

Accuracy and Reliability

Integrity Periodicity and Timeliness Openness

Utilization in Decision Making

Operational use

Accessibility Effectiveness in Disseminating Findings

Volume 14 – Issue 2– 2026

Physical infrastructure (1) Efficient use of resources (4) (2) (11) Administrative data (4) Financial data (5) Human resources data (1) Learning outcome data (5) (5) Quality assurance measures (4) Data requirement and considerations (3) System adaptability (1) Validity across data sources (15) Integrity of non-educational data bases (3) Archiving data (2) Services to EMIS data (1) Concepts and definitions (7) Classification (3) Scope (8) Basis for recording (4) Source data (12) Validation of source data (9) Statistical techniques (7) Professionalism (13) Transparency (9) Ethical standards (3) Periodicity (3) Timeliness (2) EMIS stakeholders (1) User awareness (2) User capacity (3) Utilization in evaluation (3) Utilization in governance (2) Utilization by school (5) Utilization by clients (2) Utilization by governance (2) Understandable data (8) Widely disseminated data (13) Platforms for utilization (3) User support (6) Disseminations strategy (3) Dissemination effectiveness (2)

Drawing inspiration from the SABER framework, a customized set of questions was initially developed for Motherboard users. Upon review, it became evident that certain questions were pertinent only to specific stakeholder groups. Consequently, four distinct forms were created,

209

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

each meticulously tailored with questions most suitable for the respective group of stakeholders. SABER-EMIS Tool Administration We collected data using the SABER-EMIS tool. Several steps were taken to ensure the fair and efficient administration of the SABER-EMIS tool. First, the tool was divided into 11 sections, tailored to the specific needs and roles of five participant categories: Dzongkhad Education Officers (district-level administrators responsible for overseeing educational operations), School EMIS Managers or Principals (school-based managers or leaders responsible for maintaining and using the Education Management Information System) (two sections), Policy and Planning Division Officers (officers responsible for developing and implementing education policy and planning at a national level) (three sections), GovTech Officers (technical specialists responsible for digital government solutions and infrastructure) (three sections), and Education Monitoring Officers (professionals tasked with monitoring and evaluating educational programs and outcomes) (one section). Second, these parts were then integrated into Google Forms for online administration. Rigorous quality control measures were implemented, including cross-checking the Google Forms with the original questionnaires to eliminate typographical errors. Third, a pilot test was conducted with the MIEMIS team to identify and address any potential issues. The pilot test revealed minor navigation issues in the Google Forms, which were corrected before final distribution. Following this, the Google Forms were revised and finalised. Fourth, all participants were provided comprehensive training via Zoom on completing the Google Forms. Finally, participants were invited to complete the Google Forms via email. A dedicated hotline was established to address any queries during the response period, which was set at one week. The response data were retrieved by transferring them to Microsoft Excel. Method of Data Analysis The data analysis process began with data cleaning, which was carried out through frequency analysis to ensure the accuracy and reliability of the dataset. Following this, a range of descriptive statistical techniques were applied to the cleaned data. These included calculating frequencies and generating visualisations, as well as conducting crosstabulations. The purpose of these analyses was to identify general trends within the data, undertake detailed item-byitem and level-by-level examinations, detect key contradictions, and develop actionable recommendations. Additionally, a SWOT (Strengths, Weaknesses, Opportunities, and Threats) analysis was carried out to provide a comprehensive overview. Benchmark Scale Development The benchmark levels, which are defined by a four-level scale used to establish standards for each indicator, are essential for objectively assessing progress and identifying areas for improvement within a policy area. For each indicator, these levels are categorised as latent, emerging, established, and advanced. The advanced level represents the expected standard for

210

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

each indicator and serves as the baseline for defining the other three levels. To establish the advanced level, researchers conducted a comprehensive evaluation based on general trends, item-by-item analysis, key contradictions, and recommendations, culminating in a SWOT analysis. The SWOT analysis played a pivotal role in assessing the suitability of the advanced level for each indicator. Based on the alignment of the SWOT analysis with the advanced level, the researchers collaboratively calibrated the lower levels (latent, emerging, and established). Subsequently, these calibrated levels were presented for further consultation and endorsement by EMIS stakeholders, including technical experts and managers. Following this consultative process, the final benchmarks were established. For indicators with multiple sub-indicators, benchmark levels were quantified on a numerical scale: Latent = 1, Emerging = 2, Established = 3, and Advanced = 4. The benchmark level for each main factor was calculated as the average of its sub-factor levels. Similarly, the benchmark for each of the four policy areas was determined by averaging the benchmark levels of their respective leading indicators. Finally, the overall benchmark for the EMIS was computed as the average of the benchmarks across the four policy areas. These benchmarks provided a structured framework for evaluating and monitoring EMIS development (Abdul-Hamid et al., 2017; Kendall et al., 2016). Unlike the EMIS, no benchmark or maturity scale was developed for the Motherboard because the platform is still in its developmental phase and does not yet have an established evaluation framework comparable to the SABER-EMIS maturity model. Consequently, the Motherboard was assessed using descriptive statistics, with the mean responses to cross-sectional survey items serving as indicators of stakeholders’ perceptions. This approach provided a practical basis for interpreting users’ experiences and assessing the platform’s readiness, strengths, limitations, and potential for integration with the existing EMIS. Key Findings The key findings are presented across four SABER-EMIS policy areas: Enabling Environment, System Soundness, Data Quality, and Utilization in Decision Making (Tables 7–10). The results indicate that the system is largely positioned between the emerging and established levels of development, demonstrating meaningful progress in strengthening the EMIS. Enabling Environment Policy Area The findings for the Enabling Environment policy area indicate that the system is predominantly at the emerging level, with selected components reaching the established level. The legal framework shows a mixed profile, where institutionalization is well established, reflecting strong formal structures and recognition of EMIS within the system. However, several elements such as responsibility, dynamic framework, data supply, and data sharing remain at the emerging level, suggesting that while policies exist, their implementation and coordination can be further strengthened. Notably, utilization and budget within the legal framework are rated at the latent

211

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

level, indicating limited operational use of policies and insufficient financial backing in certain areas (see Table 7). Table 7 Summary Results for the Enabling Environment Policy Area Scale (1=Latent, 2=Emerging, 3=Established, 4=Advanced)

Enabling Environment Legal Framework

x x

Institutionalization Responsibility Dynamic Framework Data Supply Comprehensive and Quality Data Data Sharing and Coordination Utilization Budget Confidentiality Organizational Structure and Institutional Process Human Resources Personnel Professional Development Infrastructural Capacity Data Collection Means Database (s) Data Management System Data Dissemination Means Budget Personnel and Professional Development Maintenance Reporting Physical Infrastructure Efficient use of Resources Data-Driven Culture

x x x x x x x x x x x x x x x x x x x x x x x x x

However, human resources remain at the emerging level, with professional development particularly weak at the latent level, highlighting a need for continuous capacity building. Infrastructural capacity presents a mixed picture: while data management systems are established, critical components such as data collection means are still latent, and others like databases and dissemination tools are emerging. Budget-related dimensions are generally at the emerging level, indicating moderate but not yet optimal resource allocation and utilization. Overall, the data-driven culture is still emerging, suggesting that while systems are in place, the consistent use of data for decision-making is not yet fully embedded.

212

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

System Soundness The findings for the System Soundness policy area indicate that the EMIS is functioning at an overall emerging level, suggesting that the system has a stable technical structure. This reflects that key system components are in place and operational, enabling effective management, processing, and use of education data. Within data architecture, the system demonstrates an emerging structure, supported by emerging data analytics and a generally dynamic system. This indicates that the EMIS is capable of handling data processes in a systematic and functional manner. However, data coverage shows a mixed pattern. While administrative data is well established, financial data, human resource data, and learning outcomes data remain at the emerging level. This suggests that the system does not yet fully integrate all key education data domains, limiting its comprehensiveness. Table 8 Summary Results for the System Soundness Policy Area Scale (1=Latent, 2=Emerging, 3=Established, 4=Advanced) System Soundness

x Data Architecture Data Coverage

x x Administrative Data Financial Data Human Resources Data Learning Outcomes Data

Data Analytics Dynamic System Quality Assurance Measures Data Requirements and Considerations System Adaptability Serviceability Validity across Data Sources Integration of Non-Education Databases into EMIS Archiving Data Services to EMIS Clients

x x x x x x x x x x x x x x

Similarly, components under dynamic system features—such as quality assurance measures, data requirements and considerations, and system adaptability—are all at the emerging level.

213

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

This indicates that although basic mechanisms exist, they are still in the process of being strengthened to ensure consistency, responsiveness, and standardization across the system. Likewise, serviceability is relatively strong, with emerging performance in validity across data sources, integration of non-education databases, archiving, and services to EMIS clients. This suggests that the system is fairly effective in delivering data services and supporting users. Overall, the findings show that while the EMIS is structurally sound and operationally effective, further improvements are needed in expanding data coverage and strengthening system adaptability to achieve a more advanced level of system soundness. Data Quality The findings for the Data Quality policy area indicate that the overall EMIS is operating at the established level, reflecting a generally strong and well-structured system. The results show that the system has a solid technical foundation, particularly in areas such as data architecture, data analytics, dynamic system functionality, and serviceability, all of which are rated at the established level. This suggests that the EMIS is capable of effectively organizing, processing, and delivering data for operational and management purposes. Table 9 Summary Results for the Data Quality Policy Area Scale (1=Latent, 2=Emerging, 3=Established, 4=Advanced) Data quality Methodological soundness Concepts and Definitions Classification Scope Basis for Recording Accuracy and Reliability Source Data Validation of Source Data Statistical Techniques Integrity Professionalism Transparency Ethical Standards Periodicity and Timeliness Periodicity Timeliness

3 x x x x

x x x x x x x x x x x x x

However, the findings also reveal uneven development across different types of data coverage. While administrative data is well established, financial data, human resource data, and learning

214

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

outcomes data remain at the emerging level. This indicates that although the system is functional, it does not yet fully integrate all critical education-related datasets in a comprehensive manner. Similarly, elements such as quality assurance measures, data requirements and considerations, and system adaptability are still at the emerging stage, suggesting that further refinement is needed to strengthen system responsiveness, standardization, and flexibility. Utilization in Decision Making The findings for the Utilization in Decision Making policy area indicate that the EMIS is functioning at an overall established level, suggesting that data generated by the system is being actively used to support planning, monitoring, and decision-making processes within the education sector. This reflects a reasonably strong culture of evidence-based decision-making, where EMIS outputs are not only produced but also applied in practice. In terms of openness, the system demonstrates an established level of performance, with strong engagement of EMIS stakeholders, user awareness, and user capacity. This indicates that relevant users are generally aware of the system and possess the necessary skills to interact with it effectively. Operational use is also well established, particularly in areas such as evaluation, governance, school-level decision-making, and government use, showing that EMIS data is embedded in core education management functions. However, utilization by clients remains at the emerging level, suggesting that external or indirect users are not yet fully benefiting from the system’s outputs. Table 10 Summary Results for the Utilization in Decision Making Policy Area Scale (1=Latent, 2=Emerging, 3=Established, 4=Advanced)

Utilization in Decision Making Openness

x x x x x x x x x

EMIS Stakeholders User Awareness User Capacity Operational Use Utilization in Evaluation Utilization in Governance Utilization by Schools Utilization by Clients Utilization by Government

x x

Accessibility Understandable Data Widely Disseminated Data Platforms for Utilization

215

x x x

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

User Support Effectiveness in Disseminating Findings Dissemination Strategy Dissemination Effectiveness

x x x x

Regarding accessibility, most components are at the established level, including understandable data, dissemination platforms, and user support. This indicates that EMIS information is generally accessible and user-friendly. Similarly, effectiveness in disseminating findings is largely established, supported by a clear dissemination strategy. However, dissemination effectiveness is only at the emerging level, suggesting that while systems for sharing information exist, their impact and reach could be further strengthened. Overall, the findings show that EMIS data is widely used in decision-making processes, but further efforts are needed to enhance broader stakeholder utilization and improve the effectiveness of dissemination practices. Overall Benchmarks The analysis identified four interconnected domains that influence Bhutan’s EMIS ecosystem. Table 11 provides a comprehensive overview of the system design. Table 11 Overall Benchmarks Scale (1=Latent, 2=Emerging, 3=Established, 4=Advanced) Enabling environment System soundness Data quality Utilization in decision making

Scale Benchmarks 1 2 3 4 x x x x

Interpreting Table 11 in the broader context of Table 3, the Enabling Environment, assessed at the Emerging level, highlights the need to reinforce governance structures, regulatory clarity, human resource capacity, and resource allocation to ensure long-term institutional ownership and accountability. Within the domain of System Soundness, key elements of system architecture and operational structures are progressing but have not yet reached full maturity. While foundational mechanisms are in place, further efforts are required to strengthen institutional coordination, improve interoperability, and enhance the system’s overall resilience and adaptability. Data Quality demonstrates comparatively stronger performance at the Established level, supported by standardized concepts, definitions, and data management processes that improve consistency and reliability, alongside regular data collection cycles and validation mechanisms that enhance accuracy and timeliness; however, continued improvements in data coverage, source data quality, and statistical processes remain necessary to sustain progress. Finally, Utilization in Decision Making is also positioned at the Established level, indicating that EMIS data are increasingly informing planning, monitoring, and

216

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

evaluation processes; improved accessibility of data platforms and growing user awareness contribute to this advancement, although broader engagement with external stakeholders and more effective dissemination strategies could further strengthen the system’s overall impact. Motherboard Findings from the situational analysis indicate that the Motherboard is a highly customised digital platform designed to support the pedagogical and assessment requirements of the Bhutan Baccalaureate. The platform accommodates substantial qualitative and quantitative learner data and incorporates secure role-based access mechanisms for different user groups. It also provides functionalities that enable learners to set goals, monitor their progress, and engage in reflective learning. Stakeholders generally expressed positive perceptions of the platform and recognised its potential to support teaching, learning, and assessment. However, they also identified several challenges, including a steep learning curve for new users, dependence on reliable internet connectivity, variability in the quality of qualitative data entered into the system, and limited accessibility for users with disabilities. Motherboard and EMIS Integration The situational analysis further revealed that the Motherboard and EMIS currently operate as complementary but largely separate systems. Stakeholders consistently identified the need for stronger interoperability to minimise duplicate data entry, improve data consistency, and facilitate more efficient access to administrative and learning information. The findings indicated that greater integration between the two platforms would enhance the availability of comprehensive educational data for schools and policymakers. Discussion The situational analysis reveals that EMIS and the Motherboard have attained a stage of functional maturity that facilitates data-driven decision-making, despite encountering structural and operational limitations that constrain their combined potential. EMIS operates at or below Established level across the SABER-EMIS policy domains, with select elements approaching Advanced maturity. Notably, strengths are particularly evident in Data Quality and Utilization in Decision Making, where core components of data architecture, administrative coverage, analytics, and dynamic system functions support routine governance and planning. Concurrently, limited integration with non-education databases remains a significant weakness, corroborating findings from the international literature that underscore interoperability as a prerequisite for policy-relevant analytics and cross-sector planning (Abdul-Hamid, 2017; van Wyk & Crouch, 2020; World Bank Group, 2011). Data quality continues to influence the extent to which both systems can facilitate effective decision-making. While methodological soundness, integrity, and ethical standards are

217

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

generally well-established, weaknesses persist in data scope, accuracy, and reliability. These gaps reflect challenges commonly reported in EMIS implementations globally, where inconsistencies in source data and validation processes undermine comparability and confidence in system outputs (Global Partnership for Education, 2019; UNESCO, 2019). In the Bhutanese context, strengthening quality assurance mechanisms is particularly crucial for addressing equity-oriented priorities, including gender equality and inclusion, which necessitate disaggregated, timely, and reliable data to guide targeted interventions (International Development Research Center, n.d.; United Nations Girls’ Education Initiative, 2010). The enabling environment for EMIS and Motherboard implementation reflects Bhutan’s sustained policy commitment to digital governance and education reform. Legal frameworks, organizational arrangements, and ICT infrastructure are largely assessed at an Established level, supported by national initiatives such as iSherig and Leverage ICT for Learning. These conditions align with evidence suggesting that EMIS effectiveness depends as much on institutional readiness and human capacity as on technical design (Abdul-Hamid, 2014). However, the findings also suggest that continued professional development and technical support are necessary to translate system capacity into consistent data use, particularly among educators and school leaders who operate at the point of data generation and application. Within this broader system, the Motherboard emerges as a complementary platform that addresses gaps inherent in largely quantitative EMIS architectures. The finding that the Motherboard is closely aligned with the pedagogical and assessment philosophy of the Bhutan Baccalaureate demonstrates its capacity to extend beyond the administrative functions of conventional EMIS. By integrating qualitative evidence with quantitative learner data, the platform supports more holistic monitoring of student learning, enabling formative assessment, reflection, and learner agency (Abu et al., 2026). These capabilities complement traditional EMIS functions, which primarily focus on administrative and statistical reporting, thereby broadening the evidence available for teaching, learning, and educational decision-making. At the same time, the Motherboard’s design introduces new challenges that mirror broader concerns in technology-intensive education systems. The challenges identified through the situational analysis suggest that successful implementation depends not only on technical functionality but also on sustained institutional support. The reported learning curve, reliance on stable internet connectivity, variability in qualitative data quality, and accessibility limitations indicate that technological innovation alone is insufficient to ensure effective adoption. Addressing these issues through ongoing professional development, improved infrastructure, user-centered design, and accessibility enhancements will be critical for longterm sustainability and equitable use (Abdul-Hamid, 2014; Abu et al., 2026). Despite these constraints, the Motherboard presents significant opportunities for innovation and system enhancement. The Motherboard has the potential to support personalised learning, targeted interventions, and professional collaboration through shared resource spaces and cross-school engagement. The prospect of integrating artificial intelligence and machine

218

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

learning aligns with global trends in education data systems that seek to move from descriptive reporting toward predictive and responsive analytics (Abdul-Hamid, 2017). These features position the Motherboard not only as a learning platform but also as a capacity-building tool for teachers, parents, and communities, extending the reach of data use beyond institutional boundaries. Data utilization across both systems is generally assessed at an Established level, with openness emerging as a particular strength. Stakeholder awareness and engagement with EMIS outputs are relatively strong, yet the operational use of data by external clients and the effectiveness of dissemination remain uneven. This pattern reflects international evidence that the value of EMIS is realized only when data circulate across institutional levels and inform routine decision-making, rather than remaining confined to reporting functions (UNICEF, 2019; World Bank Group, 2011). The labour-intensive nature of the Motherboard further limits scalability, reinforcing the case for workflow optimization, automation, and tighter integration with EMIS to reduce duplication and improve efficiency. The combined analysis of EMIS and Motherboard highlights the importance of integrating quantitative system-level data with qualitative, learner-centered evidence. While EMIS provides essential metrics on enrolment, staffing, and performance, the Motherboard captures dimensions of learning that are critical for understanding equity, inclusion, and student experience. Pilot integration in Bhutan Baccalaureate schools suggests that such complementarities are feasible, though full interoperability remains a work in progress. Advancing this integration would strengthen the evidence base for policy and practice, particularly in areas such as gender-responsive planning and inclusive education (Abu et al., 2026). Several implications arise for systems integration, policy formulation, implementation, and future interventions. In the near term, priority should be given to strengthening data validation, expanding data scope through the revamping of the existing EMIS, improving dissemination practices, and enhancing stakeholders’ data literacy and capacity for evidence-informed decision-making. Equally important is the implementation of a single sign-on (SSO) system, secure bidirectional data sharing through Application Programming Interfaces (APIs), and realtime data synchronisation between EMIS and the Motherboard. Such systems integration would reduce data silos, improve operational efficiency, and combine EMIS administrative data with the Motherboard’s learning analytics to generate more comprehensive evidence for policy formulation, planning, monitoring, resource allocation, and educational decisionmaking while maintaining appropriate data security and privacy. Over the longer term, integrating non-education databases, embedding accessibility features, automating labourintensive processes, and aligning system design with international standards would further strengthen Bhutan’s education data ecosystem. However, these developments may also introduce challenges related to technical scalability, data security, and uneven stakeholder readiness, underscoring the need for sustained capacity building, technical support, and robust governance arrangements. Collectively, these priorities reflect the broader ambition of the MIEMIS project to move beyond system functionality towards an integrated, secure, equitable,

219

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

and evidence-informed education data ecosystem that supports sustained and meaningful data use across all levels of Bhutan’s education system. Importantly, the findings of this situational analysis have already informed policy and practice. At the time of writing, the Ministry of Education and Skills Development, Royal Government of Bhutan, had utilised the findings to guide the revamping of the national EMIS. In line with the recommendations of this study, the integration of EMIS and the Motherboard is also underway, with the aim of creating an interoperable education information ecosystem that combines administrative and learning data for more comprehensive decision-making. As a key intervention under the MIEMIS project, nationwide capacity development training on the new EMIS has also been planned for all schools and relevant education stakeholders. These developments demonstrate the practical value of the research in supporting evidence-informed education reform and illustrate how situational analysis can serve as a catalyst for systems integration and continuous improvement. Conclusion This study evaluates the effectiveness of Bhutan’s EMIS and the Motherboard using the Systems Approach for Better Education Results, demonstrating that while the systems have achieved functional maturity in several areas, their overall effectiveness is constrained by gaps in enabling conditions and system integration. By synthesising the findings into a four-domain evaluation framework—enabling environment, system soundness, data quality, and data utilisation—the study provides a structured lens for understanding the performance of education data systems. The results indicate that although data quality and utilisation are approaching established levels, limitations in interoperability, capacity, and institutional alignment continue to hinder the full realisation of data-driven governance. The study highlights the importance of integrating quantitative administrative data from EMIS with qualitative learning analytics from the Motherboard to support more holistic, equitable, and context-responsive decision-making. In doing so, it advances a transferable evaluation model that can inform EMIS strengthening efforts in similar low- and middle-income contexts. Building on these insights, the paper proposes a phased roadmap that combines short-term improvements in data validation and accessibility with longer-term strategies focused on interoperability, capacity development, and international benchmarking. Bhutan’s ongoing reforms, including EMIS revitalisation and system-wide capacity building, reflect a strong policy commitment to this agenda. Collectively, these efforts position Bhutan to move beyond system functionality toward sustained and meaningful data use for improving learning outcomes.

Acknowledgements The authors gratefully acknowledge the financial support provided by the International Development Research Center (IDRC) and the Global Partnership for Education (GPE)

220

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Knowledge and Innovation Exchange (KIX) for the project, Enhancing Data Utilization and Decision-Making in Bhutan’s Education System: Motherboard Integrated Education Monitoring and Information System (MIEMIS), of which this paper forms a part. Declaration of Generative AI and AI-assisted Technologies No AI tools were consulted or utilized in the creation of this manuscript.

221

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

References Abdul-Hamid, H. (2014). What matters most for education management information systems: A framework paper (SABER Working Paper Series No. 7).World Bank. https://doi.org/10.1596/21586 Abdul-Hamid, H. (2017). Data for learning: Building a smart education data system. Word Bank Group. World Bank. https://documents1.worldbank.org/curated/en/444461505806849250/pdf/119805PUB-PUBLIC-DOCDATE-10-6-17.pdf Abdul-Hamid, H., Saraogi, N., Mintz, S. (2017). Lessons learned from World Bank education management information system operations: Portfolio review, 1998—2014. World Bank. https://openknowledge.worldbank.org/bitstream/handle/10986/26330/9781464810565 .pdf?sequence=2 Abu, M., Barkhaya, N. M. B., & Musaddad, H. A. (2026). The use of data analytics in managing education information: A conceptual review. International Journal of Business and Technopreneurship, 16(2). https://doi.org/10.58915/ijbt.v16i2.2000 Adhikari, N. P., & Budhathoki, J. K. (2025). Challenges of educational management information system: A review. ILM, 20(1). https://doi.org/10.3126/ilam.v21i1.75676 Ariko, C. O. (2024). Exploring the impact of technology on educational management information systems in Nyanza Region, Kenya: A comprehensive analysis. International Journal of Research and Innovation in Social Sciences, 8(8). https://dx.doi.org/10.47772/IJRISS.2024.8080130 Begimkulov. E., & Darr, D. (2023). Scaling strategies and mechanisms in small and medium enterprises in the agri-food sector: A systemic literature review. Front. Sustain. Food Syst.7. https://doi.org/10.3389/fsufs.2023.1169948 Bhutan Council for School Examinations and Assessment. (2021). 2021 examination highlights. https://www.bcsea.bt/ Bryson, J. M. (2004). What to do when stakeholders matter: stakeholder identification and analysis techniques. Public management review, 6(1), 21–53. https://doi.org/10.1080/14719030410001675722 Druk Gyalpo’s Institutes. (n.d). Motherboard. https://rigpa.pangbisa.com/ Global Partnership for Education. (2019). Meeting the data challenge in education. https://www.globalpartnership.org/node/document/download?file=document/file/2019 -07-15-Meeting-the-data-challenge-in-education.pdf Humphrey, A. (2005). SWOT analysis for management consulting. SRI Alumni Newsletter. SRI International, United States.

222

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

International Development Research Center. (2024). Education data systems and data use: A research synthesis. https://www.gpekix.org/sites/default/files/202502/Data%20Research%20Synthesis%20EN%20Final.pdf International Development Research Center. (n.d.). Guide to integrating gender in your project. https://idrc-crdi.ca/en/kix-call-proposals-knowledge-and-innovationstrengthened-education-data-systems-and-data-use Iyengar, R., Mahal, A. R., Aklilu, L., Sweetland, A., Karim, A., Shin, H., Aliyu, B., Park, J. E., Modi, V., Berg., M. & Pokharel, P. (2016). The use of technology for large-scale education planning and decision-making. Information Technology for Development, 22(3), 525–538. https://doi.org/10.1080/02681102.2014.940267 Kendall, J. S., Dandapani, N., & Cicchinelli, L. F. (2016). Benchmarking the state of Pohnpei's Education Management Information System. REL 2017-175. Regional Educational Laboratory Pacific. Kositsky, M. R. (2023). Education statistics: Facts about American schools. https://www.edweek.org/leadership/education-statistics-facts-about-americanschools/2019/01 Maier, C., Thatcher, J. B., Grover, V., & Dwivedi, Y. K. (2023). Cross-sectional research: A critical perspective, use cases, and recommendations for IS research. International Journal of Information Management, 70, 102625. https://doi.org/10.1016/j.ijinfomgt.2023.102625 Maqbool, R., Rashid, Y., & Ashfaq, S. (2022). Renewable energy project success: Internal versus external stakeholders’ satisfaction and influences of power‐interest matrix. Sustainable Development, 30(6), 1542–1561. Mbawala, J. J., Lestari, S., & Mwakalindile, A. (2024). The impact of educational management information systems (EMIS) on effective school management in Tanzania. Journal Penelitian Pendidikan IPA (Journal of Research in Science Education), 10(4). https://doi.org/10.29303/jppipa.v10i4.7033 Mclean, R. & Gargani, J. (2019). Scaling impact: Innovation for the public good. Routledge. McNiff, J., & Whitehead, J. (2006). Action research: Living theory. SAGE Publications. Mills, G. E. (2014). Action research: A guide for the teacher researcher (5th ed.). Pearson Ministry of Education and Skills Development. (n.d). EMIS portal. https://portal.education.gov.bt/ Ministry of Education. (2014). Bhutan education blueprint 2014-2024: Rethinking education. http://www.education.gov.bt/wp-content/downloads/publications/publication/BhutanEducation-Blueprint-2014-2024.pdf Ministry of Education. (2019). iShering-2: Education ICT master plan 2019-2023. http://www.education.gov.bt/wp-content/uploads/2021/09/iSherig-2-Education-ICTMNasterplan-2019-2023.pdf

223

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Mollerus, F., Lynch, C., & Bruining, H. (2025). Data interoperability for a systems approach to developmental conditions. Neuroscience and Biobehavioral Reviews, 176. https://doi.org/10.1016/j.neubiorev.2025.106245 National Forum on Education Statistics. (2021). Forum guide to strategies for education data collection and reporting (NFES 2021013). U.S. Department of Education. Washington, DC: National Center for Education Statistics. https://files.eric.ed.gov/fulltext/ED611949.pdf National Statistics Bureau of Bhutan. (2018). 2017 population and housing census of Bhutan. https://www.nsb.gov.bt/publications/census-report/ National Statistics Bureau. (2020). 12th five-year plan: 1 November 2018- 31 October 2023. https://www.nsb.gov.bt/12th-five-year-plan Office of the Prime Minister and Cabinet. (2023). Flagship programs. https://www.pmo.gov.bt/?stm_service=flagship-programs#1591395806422-2335c4b3028a Policy and Planning Division. (2022). Annual education statistics (34th ed.). http://www.education.gov.bt/wp-content/uploads/2023/04/AES-2022-revised-1.pdf Poudel, K. Y. (2022, December 22). EMIS problems delay school results across Bhutan. Kuensel. https://kuenselonline.com/emis-problems-delay-school-results-acrossbhutan/ Rogers, P. (2014). Theory of change: Methodological briefs, impact evaluation No. 2. UNICEF. https://www.unicefirc.org/publications/pdf/brief_2_theoryofchange_eng.pdf Stringer, E. T., Christensen, L. M., & Baldwin, S. C. (2010). Integrating teaching, learning, and action research: Enhancing instruction in the K-12 classroom. SAGE Publications, Inc., https://doi.org/10.4135/9781452274775 Tshering, G. (2025a). System soundness of EMIS in Bhutan: Situational study report. International Development Research Center, Ottawa, Canada. https://www.gpekix.org/knowledge-repository/system-soundness-emis-bhutansituationalanalysis-report-1 Tshering, G. (2025b). Quality data of EMIS in Bhutan: Situational study report. International Development Research Center, Ottawa, Canada. https://www.gpekix.org/knowledgerepository/quality-data-emis-bhutan-situationalanalysis-study-report-0 Tshering, G. (2025c). Enabling environment for EMIS in Bhutan: Situational study report. International Development Research Center, Ottawa, Canada. https://www.gpekix.org/knowledge-repository/enabling-environment-emis-bhutansituationalanalysis-study-report-1 Tshering, G. (2025d). Utilization of EMIS in decision making in Bhutan: Situational study report. International Development Research Center, Ottawa, Canada.

224

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

https://www.gpekix.org/knowledge-repository/utilization-emis-decision-makingbhutansituational-analysis-study-report-0 Tshering, G. & Bahl, A. (2023). Enhancing data utilization and decision-making in Bhutan’s education system: Motherboard Integrated Education Management Information System [Project Proposal]. International Development Research Center (IDRC). Proposal ID: 110319. https://idrc-crdi.ca/en/what-we-do/projects-wesupport/project/enhancing-data-utilization-and-decision-making-bhutans Tshering, G., & Choden, K. (2023). Analyzing test data: A mathematics test case study. RABSEL, 24(1). Tshering, G., & Phuntsho, T. (2025). Examining Grade 10 students’ ICT utilization in Bhutan. Journal of Global Education and Research, 9(2). https://www.doi.org/10.5038/2577-509X.9.2.1365 UNESCO. (2018). Re-orienting education management information systems (EMIS) towards inclusive and equitable quality education and lifelong learning. https://unesdoc.unesco.org/ark:/48223/pf0000261943 UNESCO. (2019). Education management information system. https://emis.uis.unesco.org/ UNESCO. (2022). Re-imagining the future of education management information systems: Ways forward to transform education data systems to support inclusive, quality learning for all. https://unesdoc.unesco.org/ark:/48223/pf0000381618.locale=en UNICEF. (2019). Gender toolkit: Integrating gender in programming. UNICEF Regional Office for Europe and Central Asia. https://www.unicef.org/eca/media/15101/file United Nations Girls’ Education Initiative. (2010). Equity and inclusion in education: A guide to support education sector plan preparation, revision, and appraisal. https://www.globalpartnership.org/sites/default/files/2010-04-GPE-Equity-andInclusion-Guide.pdf United States Census. (2021). Quick facts: United States. https://www.census.gov/quickfacts/fact/table/US/HSD410221 van Wyk, C. & Crouch, L. (2020). Efficiency and effectiveness in choosing and using an EMIS: Guidelines for data management and functionality in education management and information system (EMIS). UNESCO Institute of Statistics. https://emis.uis.unesco.org/wp-content/uploads/sites/5/2020/09/EMIS-Buyers-GuideEN-fin-WEB.pdf World Bank Group. (2011). Improving information systems for planning and policy dialogue: The SABER EMIS assessment tool. http://wbgfiles.worldbank.org/documents/hdn/ed/saber/supporting_doc/background/e ms/saberemis.pdf

Corresponding author: Gembo Tshering Email: gembotshering.pce@rub.edu.bt

225

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Reviewers: Volume 14 – Issue 2 The editorial team would like to thank the following reviewers for their contributions to the peer review for this issue of the journal. Their dedication and assistance are greatly appreciated. Senior Reviewers Dr Murielle El Hajj Nahas Lusail University, Qatar Email: murielle.elhajj@hotmail.com Dr Elena Mishieva Lomonosov Moscow State University, Russia Email: lenamurashkovskaya@gmail.com Leonardo Munalim University of San Jose, Philippines Email: leonardo.munalim@gmail.com Dr Shwadin Sharma California State University Monterey Bay, USA Email: ssharma@csumb.edu Reviewers Dr Rena Alasgarova Baku Oxford School, Azerbaijan Email: rena.alasgarova@mtk.edu.az Dr Aderinsola Eunice Kayode Durban University of Technology, South Africa aderini2002@gmail.com Dr Azlan Abdul Aziz Universiti Teknologi MARA Melaka, Malaysia Email: azlan225@uitm.edu.my Dr Dhritiman Chakraborty University of Bristol, UK dr.dhritiman.c@gmail.com Dr Terrence Chong UNSW Sydney, Australia terrence.chong@unsw.edu.au

226

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Dr Shiao-Wei Chu National Pingtung University, Taiwan Email: shiaowei0819@gmail.com Dr Alice Gasparini University for Foreigners of Siena, Italy Email: alice.gasparini@unistrasi.it Dr Meera Gungea Open University of Mauritius, Mauritius m.gungea@open.ac.mu Dr Hee Seon Jang Pyeongtaek University, Korea hsjang@ptu.ac.kr Dr Deepika Kohli Khalsa College of Education, India deepikakce82@gmail.com Dr Jean-Yves Le Corre Southwestern University of Finance & Economics and Shenzhen University, China jylecorre@hotmail.com Dr Yuek Li Ker Southern University College, Malaysia Email: likavia2011@hotmail.com Dr Louisa Muparuri Zimbabwe School Examinations Council (ZIMSEC), Zimbabwe lmuparuri@gmail.com Dr Suja Nair Educe Micro Research, India Email: sujarnair269@gmail.com Dr Queen Ogbomo Tennessee Technological University, USA qogbomo@tntech.edu Dr Nato Pachuashvili International Black Sea University, Georgia npachuashvili@ibsu.edu.ge

227

IAFOR Journal of Education: Technology in Education

Volume 14 – Issue 2– 2026

Dr Fei Ping Por Tunku Abdul Rahman University of Management and Technology (TAR UMT), Malaysia porchristine@gmail.com Dr G. S. Prakasha Christ University, India Email: prakasha.gs@christuniversity.in Dr Sherwin B. Sapin Laguna State Polytechnic University-Los Baños Campus, Philippines sbsapin@lspu.edu.ph Dr Faisal Syafar Universitas Negeri Makassar, Indonesia faisal.syafar@unm.ac.id Dr Miguel A. Varela International Baccalaureate (IB) School Verification/Consultant/Reader/Workshop Leader, IB Organization Email: miguevarela@gmail.com Dr Vanessa Vinodhen Nanyang Polytechnic, Singapore vanessa.vinodhen@gmail.com

228

the iafor

journal of education

ISSN: 2187-0594

IAFOR Journal of Education: Volume 14 – Issue 1 – Techology in Education