Skip to main content

Myth 11 - Herrmann Brain Dominance Instrument

Page 1


The Herrmann Brain Dominance Instrument (HBDI) measures four thinking styles located in four neat quadrants of the brain

When faced with someone who has never studied medicine or psychology, but majored in physics and music, we should be wary of any claims that they can do a better job at analyzing the brain. Such is the case with Ned Herrmann (1922–1999). Following his career as an arts performer, he worked for many years at General Electric, where his responsibility was to oversee training program design.

Key words:

HBDI, brain dominance, left-right, lateralization, preferences, childhood, triune brain myth.

What is HBDI about?

Ned Herrmann developed his “ whole brain theory ” by dividing the brain into four neat quadrants that allegedly reflect our thinking styles. Each quadrant mainly consists of “one hemisphere and one limbic half.” Each of these four parts (two hemispheres and two limbic halves) can be dominant, resulting in a preferred (dominant) thinking style. Because one of these quadrants is supposedly dominant over the others, according to this speculation, he invented the name ‘Brain Dominance Instrument’ for his questionnaire.

According to this theory, at birth our brain is “essentially whole” (whatever that means) and children develop preferred a thinking style later on as a result of life experiences. Oh yes, and this model comes with colors too, just like the DiSC and Insights Discovery. And, once again, the colors mean something entirely different!

● Blue: the analyst: logical, analytical, fact-based, quantitative thinking, situated in the upper left (neocortex).

● Green: the organizer: organized, sequential, planned and detailed thinking, situated in the lower left (limbic system).

● Red: the communicator: interpersonal, feelings-based, kinesthetic and emotional, situated in the lower right (limbic system).

● Yellow: the creator: holistic, intuitive, integrating and synthesizing, situated in the upper right (neocortex).

To assess these thinking styles, a questionnaire consisting of 120 items (current) or 160 items is administered to participants. Participants in HBDI courses receive beautiful, multicolored reports. The report details 9 main scores: left and right brain dominance, four quadrant scores, ‘cerebral and limbic’ preferences, and introversion versus extraversion. According to the HBDI Group, one of the major applications is the link between preferred thinking styles and jobs that require that particular thinking style. It can therefore help you detect your vocational interests. But it can also help you acquire a deeper understanding of your communication styles, problem solving, and preferred learning. You can learn to achieve better sales, to think creatively, and manage change better in order to transform

your entire business! The Group’s ‘experience’ (not ‘research’) has shown that the applications are “almost without limit.”121 Strong claims, it looks like the Holy Grail.

Nowadays, Herrmann International claims that everyone can be trained to flex their thinking styles and that managers, for example, will make better decisions if they learn to use all four brain quadrants:

“Although you may prefer certain modes of thinking over others, everyone has access to all four quadrants. Using the skill of Whole Brain Thinking, agile leaders are able to fully leverage their preferences, stretch to other styles when necessary, and adapt to take an advantage of the preferences of those around them to be faster, more focused and more flexible in any environment.”122

Let’s see whether these strong claims are backed up by strong evidence, the very least we can expect in this case.

■ Executive Summary Theory

The HBDI theory posits that the brain can be divided into four neatly distinguishable brain regions (a combination of “left-right hemisphere” and “upper-lower” regions) that produce four neatly distinguishable thinking styles. Herrmann “integrated his concepts with Left Brain/Right Brain and Triune Brain theories into a new “brain dominance theory,” and assumed that we develop a preference for one thinking style during early childhood. The theory was entirely wrong, as the current status of brain research tells us a totally different story. Lateralization in left-right hemispheres is limited to a number of functions: our modular brain mainly works in networks, distributed throughout the brain. The HBDI theory is to be considered a highly oversimplified representation of early split-brain research.

Ned Herrmann abandoned his idea long ago, though the pages on the HBDI website don’t reflect this change of opinion. Indeed, he later admitted: “The whole-brain model, although originally thought of as a physiological map, is today entirely a metaphor” (1989, p. 63).

Empirical data

● The empirical data on brain lateralization (left-right) tells a much more complex story, leaning more in the direction of distributed functional neural networks.

● The data don’t allow for a conclusion to be made about dominant brain regions.

● There are no peer-reviewed published studies confirming the beneficial effects (learning to adapt to other thinking styles, thinking faster, being more focused, more flexible, enhanced innovation and revenue growth, increased productivity and engagement, etc.) of training based on the HBDI concept. Strong claims with a total absence of evidence.

● Although Bunderson wrote in 1998 that the supposed four quadrants in the brain should be backed up by actual brain research, no such research has been conducted by HBDI since 1989. Obviously, it is impossible to determine from just a questionnaire what brain regions are involved in thinking.

121 Website www.hbdi.com (now thinkherrmann.com): The Theory Behind The HBDI and Whole Brain Technology (text from December 1, 1998).

122 I cite this text from the ‘White Paper’ (page 7) from Herrmann International: “Navigating in an Unpredictable and Complex World – Why Thinking Agility is Critical to a Manager’s Success.”

Test construction

● I could only find two internal documents (1989 and 2014), both of which were written by C. Bunderson, a psychologist who claims to be specialized in psychometrics.

● The test construction doesn’t offer any evidence of either brain lateralization, the neat division into four regions, or the alleged brain dominance. The choice of words ‘left’ and ‘right’ for the two factors the authors of the questionnaire selected, is an entirely arbitrary choice. The assumed bipolar nature has not been demonstrated, in neither the first nor the second pa-per. The reported correlations between the “four quadrants” differ from those expected mathematically (geographic location by calculating cosine).

● Based on these internal papers, I will discuss several problems, including the lack of clarity regarding the number of factors found, how much vari-ance each factor explains, the effects of a forced choice (or ipsative) format with only four scales, the use of forced scores as if they were continuous scores (an inadmissible mathematical calculation), the omission of certain parts of the questionnaire before statistical analysis, etc. Some of the sta-tistical methods used differ quite a bit from standard test construction methodology. In at least one instance, unacceptable mathematical calcula-tions are made using ipsative scores. This makes it impossible to assess the true reliability and validity of the questionnaire.

● That being said, I want to remind the reader that it is possible to construct a questionnaire with seemingly good psychometric properties, even if the theory is entirely wrong. It reminds me of the dead salmon hoax described in the brain scan myth.

Theoretical/empirical grid

Conclusion

The concept of thinking styles being located in four quadrants of the brain is a ludicrous theory, and attempts to prove that the questionnaire is a valid and reliable measure of such brain dominance and thinking preferences have failed.

Moral Assessment

If even the founder (Ned Herrmann) has admitted the model is wrong, and if research has demonstrated that the brain works in a totally different manner than Herrmann had assumed, one can only conclude that proponents are making good money with the HBDI thanks to so many gullible HR people and managers. Like other models, it is unacceptable to use tools that might negatively impact career choices and decisions, have a negative economic impact or change one’s self-image.

Discussion

■ Theoretical soundness

The HBDI writes that Ned Herrmann combined the two brain halves theory that he attributed to Roger Sperry with Paul MacLean’s triune brain theory. By combining the left-right division and the horizontal distribution in three parts (triune), he obtained four quadrants. Eureka! Roger Sperry, however, never described the left-right division as a general division, he only noted some effects on epileptic patients by cutting the corpus callosum, thus reducing epileptic attacks in patients, while noticing some other effects. Paul MacLean hypothesized that our brain consists of three parts: a reptilian brain, a limbic or ‘paleomammalian brain,’ and the neocortex or ‘neomammalian complex.’

Since no peer-reviewed evidence for the four quadrants in the brain, as assumed by Ned Herrmann, can be found in the academic database, we must rely solely on what HBDI proponents and believers tell us. From the Bunderson ‘validation document,’ which anyone can download from the HBDI website, we learn the following assumptions: the preference for a thinking style follows the preference for right-versus left-handedness. Second, people prefer “one or two clusters” of information processing over other clusters: this is referred to as ‘dominance’—expressing the dominant preference. It is further speculated people may avoid using other clusters (one or two). This remains pure speculation and is not substantiated by any brain research evidence presented in the Bunderson paper.

They answer their own question “ Is the fourfold or quadrant model of brain dominance supported by research data? ” in the following manner: “ Four stable, discrete clusters of preference exist.” A careful reader will notice this doesn’t answer the question of whether brain dominance is supported by research data. The answer is rather a reply to another question: ‘Are there four preferences?’ These preferences are of course the result of a statistical analysis of the item-scores.

The current status of the sciences tells us a very different story than the one propagated by Herrmann and his successors. The brain cannot be divided into four separate quadrants as recent brain research has abundantly demonstrated. Put simply, the neat division of our brain in 4 quadrants is entirely false.

This is how it actually works: our brain consists of functional networks and only some of those networks have a fixed spatial location. Most of the other functional networks are actually clusters of networks that are spread out over the brain and interact. I will discuss this extensively in Chapter 1 of Part V.

The curious case of… a metaphorical brain

In the research manual, the HBDI model is often referred to as ‘metaphorical.’ Indeed, the website for Herrmann International Europe states: “Herrmann’s framework and HBDI® are built on a metaphoric representation of the brain, which identifies four quadrants, A, B, C

and D.” Using a metaphor to describe the brain? Strange, because they also refer to brain scientists who map the brain in a ‘real’ way and not a metaphorical one. I suspect this is just another example of a switcheroo: in case you can no longer defend your claims about the left-right dichotomy (or dominance, or one quadrant dominating the others) against criticism from serious brain researchers, then just call it a metaphor.

By the way, did you know that in the initial questionnaire your level of extraversion used to be measured by… only one question? How’s that for construct validity? Reliable questionnaires such as the NEO-PI-R use 44 questions to measure extraversion. And did you realize that your dominant brain quadrant can be identified by how you hold a pencil? Should it be any surprise then that Herrmann International was awarded the Scam of the minute award by Skepdic.com (April 6, 2008). This is what Skepdic wrote:

“You might call the HBDI Myers-Briggs for Dummies. Instead of sixteen basic types, the HBDI has only four, based on four alleged quadrants of the brain, one of which is said to be the dominant form of thinking in each of us. One type is creative (the imaginative person); another is analytical and logical (the rational person); another loves order and is good at following directions (the organized person); the last type is more emotional and social (the interpersonal person). This instrument is used by corporations, governments, educational facilities, and independent consultants to determine such things as who gets hired, fired, promoted, or assigned to which task. The word on the street is that it works just as well as using astrological signs or graphology.”123

In the FAQ section on the Herrmann International website, we can still read what Ned Herrmann thought about astrology (which is total pseudoscience, of course):

“I think of astrology as a very complicated construct based on ancient myth and anecdotal data. Even after its existence for well over 100 years it remains highly speculative and scientifically invalidated. One of the reasons that the birth date question is included in the research questions portion of the survey form is to attempt to discover the degree to which there are brain dominance correlations with astrological signs.”

Lastly, you may be interested to learn that Herrmann International also helped spread Brain Gym, a method described by Ben Goldacre124 as “riddled with transparent, shameful and embarrassing nonsense.” One of the exercises is to “enhance the experience of whole brain learning.” It was very easy for Ben Goldacre, a medical doctor, to debunk Brain Gym as it spreads entirely false claims such as:

● “Processed foods do not contain water.”

● Brain Gym say you should not drink soup because “All other liquids are processed in the body as food, and do not serve the body’s water needs” Their support in spreading Brain Gym as a bad meme was one component that led them ‘deserve’ the Scam of the minute award in 2008.

However, you may want to participate in an HBDI session, because after all, in many countries you receive a present: a little bear in the color of your thinking style!

A little bit of Sherlock Ned Herrmann claimed that several notable scientists contributed to his model. But is this true? Let’s first look at the claims. In the frequently asked questions section of the website (www.hbdi.com, 2005), we find the following Q&A:

123 http://skepdic.com/news/newsletter90.html

124 Ben Goldacre (2008). Bad Science And on the internet: http://www.theguardian.com/commentisfree/2011/may/28/bad-science-goldacre-brain-gym

“Q: Is the brain dominance theory and the whole brain concept valid? What proof exists?

A: The brain dominance concept has been strongly validated in a number of different ways; First, through the research and experimentation of leaders in the field including Roger Sperry, Robert Ornstein, Henry Mintzberg, and Michael Gazzanniga. Secondly, it has been validated by the hundreds of EEG experiments carried out personally by Ned Herrmann. Third, it has been further validated by the public demonstrations conducted by Ned Herrmann over the past 15 years. Fourth, it has been validated by specific validation studies carried out by C. Victor Bunderson and James Olsen of Wicat and later by C. Victor Bunderson and Kevin Ho, and in parallel with those studies by validation experiments carried out by Schadty and Potvin at the University of Texas. Additional validation comes from the more than 60 doctoral dissertations based on both the HBDI and the whole brain concept.”

We can also read an even more explicit claim:

“Q: Who are the key contributors to Whole Brain Technology™?

A: Roger Sperry, Robert Ornstein, Henry Mintzberg, Paul MacClean, Joseph Bogen, and Victor Bunderson.”

Some of these people are still alive, so I decided to contact them. Henry Mintzberg is not a brain researcher, but an organizational psychologist. This is what he wrote to me on January 12, 2005:

“Dear Mr. Vermeren

I certainly never worked with Herrmann nor do I recall citing his work, and I did no research as such on the brain, just an article using Sperry and Ornstein’s work called “Planning on the Right Side and Managing on the Right.”

Thank you for this.

Henry Mintzberg”

Roger Sperry was a celebrated Nobel Prize-winning brain researcher, so this reference is quite significant. To check the claims about his ‘contribution,’ I contacted his research team. This is what I asked M.D. Antonio Puente back in December 27, 2004:

“Question: Has Roger Sperry ever worked with Ned Herrmann?

Question: Did Roger Sperry ever support the theory or tests from the late Ned Herrmann?”

Dr. Antonio E. Puente wrote back:

“I am not familiar with Ned Hermann or the HBDI. I do not believe that Dr. Roger Sperry was aware of either the individual nor of the instrument.

My impression, having reviewed the website, is that this system is a business consulting one based very loosely on recent theories of brain functioning, personality typology, and industrial-organizational psychological principles.

I will forward this to Mrs. Sperry to determine whether my answers are valid.

Thank you,

Antonio E. Puente”

Part two followed on December 30, 2004:

“Have now heard from Mrs. Sperry and Dr. Arches; Mrs. Sperry- Roger nor her had / have ever heard of the concept nor of the person

Dr. Bogen- “Ho, Ho, Ho” (he thought it was a practical joke on my part)

Oh, well... Happy New Year.

Tony Puente “

So, neither Roger Sperry nor Joseph Bogen, who are real brain researchers, have ever collaborated with Ned Herrmann. Henry Mintzberg never worked with Herrmann either. It seems Ned Herrmann had a vivid imagination and wasn’t worried about the risk of such false claims one day being unveiled. And Victor Bunderson? Well, he isn’t a brain scientist. According to the HBDI website, he has a “ background in educational measurement and psychometric methods.” Elsewhere, I found out he received his Bachelor of Science degree in Psychology from the University of Utah and received his Doctorate from Princeton University.

What else is wrong with this stuf?

I first published articles and a book on this subject back in 2006, and to my pleasant surprise, HBDI has not changed many of the texts, though they were criticized quite a lot at the time. Most likely the Dutch review, or reviews on skeptic websites, didn’t do much damage, so they carried on with their business as usual.

Take this example from their current website:

“Upon superficial examination, the brain consists of two half brains and therefore the simplistic left brain/right brain dichotomy seems to adequately describe the differences between these two brain halves. Closer scrutiny of the brain reveals four structures rather than just two. These four paired structures consist of the two cerebral hemispheres and the two halves of the limbic system. These are paired structures connected by what are called commissures. The two cerebral hemispheres are connected together by the Corpus Callosum and the two halves of the limbic system are connected together by the Hippocampal Commissure.”

“Extensive data speaks loudly that there are an equal number of people whose mental preferences are primarily cerebral or limbic as those that are primarily left or right.”

These explanations fly in the face of everything that true brain experts can tell you. The explanation sounds ‘science-y,’ but is far from scientific. These wild speculations about our preferences have never been proven.

I found several other claims on the Herrmann websites that contradict the current state of science. For example:

“While there will always be exceptions, we feel that on the average nurture represents 70% of who we are compared to 30% for nature. It is the profound influence of parenting, schooling, friendships, work, and life experiences in general that shape our mental preferences that in turn lead us to do what we do the way we do it.”

Perhaps ‘feeling’ is not the right way to address brain and personality science. The current scientific position has demonstrated (through twin research) that the genetic component alone accounts for 40 to 69% of variance in personality traits.125 My Champions League no longer talks about nurture versus nature, but rather about how personality is the result of inherited traits (through genes) and other biological factors (e.g. hormonal influences on the fetus during pregnancy) and learning processes (interaction with the environment).

What does my Champions League of experts say?

Neuroscientist John Bruer put it like this:

125 See the discussion in the chapter on Recruitment in Part V of this book.

“Right brain versus left brain, or hemisphericity, became a popular and influential theme in brain-and-education circles through the mid-1980s. It persists in the popular mythology and in educational circles, despite neuroscientist’s criticisms that educational applications of hemisphericity are based on a highly oversimplified understanding of the underlying research.”126

Take that.

Both Joseph Ledoux and Antonio Damasio, considered by many to be the foremost expert brain researchers, have entirely refuted the ideas of the both triune brain (MacLean) and the simple left-right division of the brain that results in hemispheric dominance. Michael Gazzaniga, a colleague of Joseph Bogen and Roger Sperry who worked on ‘split-brain patients’ obviously concurs: in an interview by Tony Retrosi on April 17, 2015, he dispelled they myth of the hemispheric brain dominance calling it “overly simplified and overstated.” Moreover, these days it is widely accepted that no decision can be made without also involving emotion.

What does the majority of the feld of experts think?

The field of experts follows the same line of thinking as my Champions League. The left brain versus right brain division is considered a neuromyth by all knowledgeable experts. The other important cornerstone of the HBDI theory is highly problematic: MacLean’s triune brain hypothesis (1960) has been “out of favor for a couple of decades or more” according to neuroscientists like Bill Skaggs (Quora, November 28, 2014). All vertebrates have a reptilian core and it is important to select actions at all levels of the brain. The limbic system is far from the sole responsible area for processing emotions. The cortex has essentially the same structure in all mammals (which have six layers in their neocortex, more than non-mammalian vertebrates). I could go on and on, but the triune brain is only a footnote in the history of brain speculation.

Jeff Anderson summarizes the myth of left brain versus right brain thinking applied to personality pretty neatly: “ The neuroscience community has never accepted the idea of ‘left-dominant’ or ‘right-dominant’ personality types. Lesion studies don’t support it, and the truth is that it would be highly inefficient for one half of the brain to consistently be more active than the other” (interview in The Guardian, November 16, 2013). In a recent study, he and other coauthors write: “Lateralization of brain connections appears to be a local rather than global property of brain networks, and our data are not consistent with a whole-brain phenotype of greater ‘‘left-brained’’ or greater ‘‘right-brained’’ network strength across individuals” (Nielsen et al., 2013). They further summarize the research by saying that only small structural asymmetries are present in most individuals. The ‘whole’ brain, or the left brain versus right brain dominance, is a myth, and so are lateralization differences between the sexes.127

The theoretical score: -5. Because the theory has been refuted, yet its proponents still try to make the impression that it is true science.

■ Empirical findings

What is the level of evidence?

Neuroscientific findings have now firmly refuted the concept of a global brain difference, global hemispheric brain dominance, or that one hemisphere is favored over another (e.g. Nielsen et al, 2013)

126 John T. Bruer (1999) The Myth of the First Three Years. (p. 35)

127 You can download and consult the article for free from Plos one (http://journals.plos.org/plosone/ article?id=10.1371/journal.pone.0071275 )

Specializations in our brain occur in either distributed neural networks or local neural networks, mostly situated in lobes and their substructures. For example, we can distinguish between the temporal lobe (e.g. sound perception, smell, learning, language), the parietal lobe (integrating information from different senses, spatial orientation, number representation…), or the frontal lobe (decision-making, planning, problem solving, thinking …). Each lobe consists of several substructures. For example, the frontal lobe consists of many substructures such as the prefrontal cortex, the orbitofrontal cortex, the motor and premotor cortex, etc. More importantly however, most complex skills, such as making calculations or word recognition, are the result of coordinated action between specialized neural networks in different parts of the brain.

So, now that we know that the theoretical foundations of the four-quadrant brain and the alleged dominance or preference are wrong, we should in fact stop any further investigation into the reliability or the validity of the questionnaire. Many people ignore the fact that one can build a questionnaire with seemingly good psychometric properties even if the theory behind it is dead wrong or the constructs are ill-defined. But I hypothesized that I would find the same sloppy research as in the theorizing, so I put my efforts in evaluating the questionnaire too..

The HBDI claims its questionnaire is valid and reliable. After all, Victor Bunderson has a PhD in psychology. This should lend him credibility even though he also believes in other myths: learning styles.128 Let’s have a look.

The first obvious problem is the lack of academic peer-reviewed publications about the validity and reliability of the HBDI questionnaire. But wait a minute, what about the “extensive data” that, according to the HBDI website, prove that the questionnaire is valid. These data are nowhere to be found in the peer-reviewed databases. At best there is a list of references to all kinds of dissertations that have (blindly) used the HBDI. Indeed, they used it without first assessing its reliability and validity. That’s the level of doctoral research sometimes. My search for peer-reviewed articles129 provided the following information:

● APA PsycNET: “HBDI AND Peer-Reviewed Journals only - did not match any documents.”

● APA PsycNET: Herrmann and Peer-Reviewed Journals resulted in one match in the Journal of Psychological Type, a magazine dedicated to… typologies such as Jungian typology. We should disregard it, as it is not a validation study on psychometric reliability and validity. Nor is comparing two bad tests very useful.

● ProQuest (ABI/INFORM Global): “Your search for ti(HBDI) found 0 results.” (note: ti = title)

● ProQuest (ABI/INFORM Global): using HBDI ‘anywhere’ & ‘peer-reviewed only,’ 16 matches with the Herrmann HBDI were found. None of these articles deal with validation studies. Most articles merely mention the existence of the instrument, some have used the instrument (e.g. Clayton & Kimbrell, 2007; Van Oordt et al., 2014130)

● EBSCO (Business Source Elite): using HBDI and ‘anywhere,’ limited to ‘academic journals’ gives two matches. The Rowe et al. (1992) paper merely assumes HBDI indeed “assesses which quadrant of the brain is most active.” The Deshpande & Baxis (2011)

128 In a document found on the website of his EduMetrics Institute we can read: “The GELC model for education, elaborated by Dr. C. Victor Bunderson, uses computers and technology to deliver a highly-customized lesson to each individual, based on that person’s learning preferences and learning style.”

129 Search conducted on October 21, 2015.

130 These are published in the non-psychological magazines Managerial Finance and Meditari Accountancy Research (MEDAR), respectively.

paper mentions use of the HBDI and a high “acceptance of their thinking preferences as well as avoidances” by the managers. The Forer effect, of course.

The ‘extensive data’ turned out to be a ‘total absence’ of peer-reviewed validation studies. Needless to say, I found absolutely no evidence in peer-reviewed papers about the strong claims such as ‘more agile thinking,’ ‘better sales,’ ‘being faster or more focused,’ etc. Strong claims, but a total lack of evidence. So again, we might as well stop here, given that we have a problematic theory and a lack of independent peer-reviewed articles.

To assess the psychometric properties of the HBDI questionnaire, we must rely on the internal HBDI papers and assume that no data were fabricated (fabricating data is a practice that even some scientists do, as we have found out in recent years131) and that they really conducted the research. Considering their fabrications regarding the existence of a quadrant brain, I find it very hard to believe them.

The first document I will discuss is the 1989 paper by Victor Bunderson (in fact a kind of summary of a book they published). Strikingly, the document doesn’t deliver any kind of evidence for the neat separation of the brain into four quadrants. At the end of the paper, Bunderson even suggests there is a need to conduct this research. Indeed, the only way to prove the four-quadrant structure would be to conduct research using real brain research techniques (e.g. using fMRI or other more recent techniques such as ‘resting state functional connectivity MRI’ or rs-fcMRI). This is typically done in the following manner: neuroscientists calculate average ‘brain images’ of large groups of healthy people. This can provide a rough idea of the functional specification of the neural networks in the brain. Using a questionnaire, as the HBDI does, just cannot offer any proof of where such thinking preferences would be spatially located in the brain.

Even 1989, Bunderson already somehow seemed to know that something was wrong, as he referred to the constructs in a ‘metaphorical’ sense (p. 6):

“The constructs measured by the HBDI are summarized metaphorically in reference to Ned Herrmann’s insightful image of the rotated triune brain, seen from the back of the head, with emphasis on the dual cerebral and limbic hemispheres. Figure A-1 is a redrawing of Figure 3-12.”

A D

B C

Figure III.11: figure adapted from the 1989 Bunderson document (p. 6).

131 One of the recent cases was that of Dutch social psychologist Diederik Stapel (over 50 papers retracted), but there have been many others such as Hisashi Moriguchi, William McBride, Andrew Wakefield, etc.

The report contains a text that seems to acknowledge the problematic and oversimplified division of our brain:

“Figure A- 1 suggests geographic location of function. Thus, it is perhaps the source of some criticism.” (Bunderson, 1989, p. 6)

Perhaps? Some criticism? As I mentioned in the theoretical section, the academic literature tells us something entirely different, so ‘some’ should be replaced by ‘overwhelming’ and ‘fundamental.’ I know I am repeating myself, but the current status of empirical research findings is negative in regards to (1) the alleged ‘four quadrants’ consisting of the hemispheres and ‘half of the limbic system,’ and (2) the triune brain by Maclean. Joseph Ledoux clearly relegated MacLean’s triune brain to the dustbin (The Emotional Brain, p. 92-103), and pointed out that reptiles do have a limbic system and even a very small ‘kind of’ neocortex.

In the 1989 Bunderson ‘validation paper,’ we also witness the author’s following effort to protect himself against criticism:

“The HBDI was not validated for use in clinical or diagnostic testing, nor in medical or psychological classification. It was not validated for use in admissions testing prior to educational or training programs nor for placement at different levels within these programs.

(Using a test score as a condition for admission is an application distinctly different from the use of the HBDI to assemble optimal seminar groups.) It was not validated for use in selection testing for employment, for professional and occupational licensure and certification, nor for making a decision about a person that is beyond the control of that person.” (p. 3)

So basically, where can it be of use? The long list of situations where the HBDI shouldn’t be used starkly contrasts the claims found on the Herrmann websites.

The stats and the psychometrics

At the very least, could the HBDI questionnaire demonstrate some minimum psychometric requirements? What do you think? Bunderson states:

“Since the Herrmann Brain Dominance Instrument is designed to measure preferences, and is not a test of knowledge, content validity is not an important issue here.” (1989, p. 4)

Not important? Let’s analyze that. To assess whether a test or questionnaire is valid and reliable, a great deal of tedious work must be done. Below I have inserted a table to compare the academic standards with the actual practice reported by Bunderson:

Academic terminology and standards

Statistical reliability: does the test or questionnaire measure adequately, without too much error margin. Can the measurement be repeated and provide the same outcome? It is NOT proof of the theory.

● sub: internal consistency (internal reliability of scales): Here we measure all correlations between the items used in a construct, concept, or scale. The most commonly used method is to calculate Cronbach’s alpha. Minimum is set at 0.70 – 0.80, which is already very high.

● sub: test-retest reliability: does the test or questionnaire provide the same results when administered x weeks later?

● sub: split-half reliability: the test is split into two even parts: the results should correlate strongly.

● sub: parallel reliability: two tests are developed that should measure the same concept. The outcomes should show high correlations with each other.

Construct or concept validity: does it really measure what the theory says. A validity coefficient can never be higher than the reliability coefficient.

● sub: convergent validity: is it related to other measures with which it is theoretically predicted to correlate (But not measure the same construct: this is concurrent validity).

● sub: congruent validity: do two measurement instruments measure the exact same construct.

● sub: discriminant validity: are the constructs (or scales) sufficiently differentiated (they should be as unrelated as possible, expressed in no or very low correlations) or different from other constructs (not covered by the questionnaires).

HBDI results from the Bunderson 1989 paper

A table with internal consistency coefficients is not provided.

The document states: “no empirical data on test-retest stability were systematically collected.” There is a reference to Kevin Ho who found ‘”78 repeated measures of the same persons” in a large database and calculated reliabilities of .91, .96, .97 and even .98. This is not proper procedure for conducting test-retest research, and the very high reliabilities are very uncommon and should be doubted. 0.98 is an almost perfect measure and should ring some alarm bells.

The HBDI is compared to the MBTI, without explaining why the HBDI (measures thinking styles) would relate to the MBTI (measures personality). They do not explain why they would theoretically correlate.The HBDI is compared to many other mostly unknown tests. Thus, huge datasets are created, offering the opportunity to detect many correlations, which of course increases the risk of these being totalaly random or spurious. To avoid dustbowl empiricism, correlations should have been predicted beforehand using a guiding theory.I also found a very unusual way (underlined) to determine whether or not to include a test: “Two tests, Necessary Arithmetic Operations Test and the “Similarities” subtest from the Wechsler Adult Intelligent Scale (1958) were chosen after telephone communication with neuropsychologists indicated those as good markers for cerebral left processing” (p. 11). Of course, such claims are impossible to verify, but it seems to me that neuropsychologists would have been familiar with the left-brain myth.

Academic

terminology and standards HBDI results from the Bunderson 1989 paper

Content validity: this is a non-statistical type of validity: experts should systematically examine whether the items of the test cover a representative sample of the domain to be measured. Scope: the items should fully cover the construct.

● sub: representation validity: is the construct operationalized (‘translated’) into observable measures.

● sub: face validity: is NOT considered to be validity: it has to do with whether people recognize something “at first sight.” It is often called false/faith validity. There is always the risk of the Forer effect.

Criterion validity: this is an indication of whether the measure is consistent with what is already known.

“Since the Herrmann Brain Dominance Instrument is designed to measure preferences, and is not a test of knowledge, content validity is not an important issue here” (p. 4). Although acknowledging ‘face validity’ does not belong to the “latest Standards for Educational and Psychological Testing –1985,” an entire page is devoted to it..

“Future Construct Validation Research: Criterion-related evidence of the reliability and validity of these predictable relationships with different profiles could be readily obtained, extending the research base for new valid applications of the HBDI” (p. 27).

● sub: concurrent validity: is it related to other established measures that measure the same theoretical concepts. A correlation of 0.50 is the minimum to be expected.

● sub: predictive validity: can the test predict something about what one wanted to know (i.e. criterion).

o sub: incremental validity: not really considered real validity. It has to do with whether a test can add extra information or extra predictive power on top of another test or questionnaire.

Internal validity: is the research design conducted with care and rigor? Were alternative explanations explored? Have other variables been considered?

External validity: Are the results of the study generalizable (e.g. to other populations)? What about possible bias?

Ecological validity: is the research setting representative of the real world? Will results obtained in a lab also be produced in the real world?

No comparison was made with established tests that purportedly measure ‘thinking’ styles.I

The study does not report predictive validity studies.

All studies and factor analysis were conducted starting from the premise that they would find the constructs.

Not reported.

Not reported.

I One such instrument used for cross-validation in this paper is the Hill ‘Cognitive Style Mapping Instrument,’ but this only appeared in a book and is not considered a valid test by the APA.

A careful reader will notice that the choice of tests for convergent validity seems arbitrary at best, while concurrent validity was not tested against established tests. Moreover, both Jungian typologies and learning styles theories have been refuted (see elsewhere in this book), so it seems they just performed random test comparisons. I can only speculate about their reasoning, but one hypothesis is that they did so with the goal of finding positive correlations that could be explained afterwards (a case of dustbowl empiricism or circular reasoning).

A remarkable inconsistency can be found in the 1989 paper. If our brain structure can be represented in four equal parts, as depicted in the circular & quadrant image below, then the correlations between the midpoint of a quadrant (near where the letters are positioned) should theoretically be the following:

● quadrants A and C should show a negative correlation of -1 as they are presented as opposites. The midpoint of quadrant A and C show a geographical relation of 180°, and math teaches us that cosine equals the correlation of two constructs, so cos(180) = -1;

● quadrants D and B: the same as above, a negative correlation of -1 should be shown;

● adjacent quadrants should be unrelated (A with B, A with D, D with C, and so on) or in other words, have a correlation of 0 because cos(90) = 0.

A D

LOGICAL ANALYTICAL FACT BASED QUANTITATIVE ORGANIZED SEQUENTIAL PLANNED DETAILED

HOLISTIC INTUITIVE INTEGRATING SYNTHESIZING INTERPERSONAL FEELING BASED KINESTHETIC EMOTIONAL

B C

This is not what is reported in the paper however:

● A-D (expected correlation = cos(90) = 0): -0.53 (Bunderson himself writes: “-.53 seems too large in absolute value,” p. 24).

● A-B (expected correlation = cos(90) = 0): 0.08 (Bunderson himself writes: “.08 seems too low”—which is strange, as it is close to what is mathematically expected—p. 24).

● A-C (expected correlation -1): -0.68.

● D-B (expected correlation -1): -0.77.

In conclusion, the circular representation is not substantiated by these correlations. Moreover, the whole summary paper is quite chaotic and unclear and uses unscientific language (e.g. “ the constructs measured by the HBDI are summarized metaphorically in reference to Ned Herrmann’s insightful image of the rotated triune brain”—p. 6) and anecdotal information (such as how happy Ned Herrmann was when some of the data were presented to him).

In 2014, 25 five years after the first publication, Bunderson and two of his colleagues from the organization he founded published a ‘final draft’ of a second validation study. From this paper we learn that the brain research recommended by Bunderson in his 1989 paper was never carried out. Thus, we still have no research (let alone evidence) into the supposed brain division into four quadrants. Twenty-five years seems more than enough time to do this, particularly since extraordinary claims require extraordinary evidence. Again, the only research conducted at this time is whether a series of questions shows some factors or clusters. The paper tries to convince us that there exists an indication of a ‘left’ and ‘right’ factor and a four-quadrant structure. Even if we were to accept that the questions could demonstrate a two-factor solution or a four-quadrant structure (as a plot in their Figure 4a indicates), it isn’t a convincing argument to call them ‘left’ (hemisphere) or ‘right’ hemisphere or ‘upper left,’ ‘lower left,’ ‘upper right,’ and ‘lower right’ (brain regions). The naming of these factors is an arbitrary and self-serving way to make a link to the alleged brain quadrants. Looking at the questions, I have no clue as to how they could offer evidence for the underlying structure in the brain. Why not call them ‘analytical,’ ‘mystical,’ ‘intuitive,’ or ‘magical’?

This new study includes (high) internal consistency reliability scores and a series of statistical calculations that just might sweep you off your feet,132 or discourage you from reading it. The paper doesn’t mention that the questionnaire uses a forced choice format—an extremely important omission: we have to know whether the reliabilities and independencies of the scales are not purely artificial due to an ipsative scoring format (remember the Clemans formula I explained in the Belbin chapter; with this formula you can ‘predict’ what intercorrelations based on ipsative scores will look like just by taking into account the number of constructs or scales used).

I had to do some Sherlock work again to find out whether the questionnaires used a normative or an ipsative format. I found an HBDI questionnaire used in a dissertation that contains 160 items and forces you to choose exactly 40 alternatives:

● 5 alternatives from a list of 20 statements. They present a forced choice between 20 items five times;

● 4 alternatives from a list of 16 items. Two such lists are presented;

● 1 alternative from a list of 4 items. They do this 4 times;

● 3 alternatives from a list of 12 items.

Each time the ratio is exactly 1:4, which clearly indicates that they are trying to find four constructs.133 Some items are highly questionable, and some make me wonder whether people would really understand the wording (e.g. “I see the ‘facts’ treating them intuitively and holistically”; Schuler, 2014, p. 60). Moreover, some sentences include multiple items, meaning it isn’t clear as to what ‘part’ or concept of the sentence you will answer (e.g. “I look for a vision that is intuitive, conceptual, visual and with ‘general context’” includes three different concepts: intuitive, conceptual, and visual; Schuler, 2014, p. 60).

We’ve learned that the questionnaire is entirely ipsatized (called “preference choice under scarcity” by Bunderson (2014, p. 6), and by now the reader will know that ipsatized formats will result in (1) artificial bipolar constructs and (2) artificial reliability coefficients (as demonstrated by an entirely artificial intercorrelation matrix). But to investigate the psy-

132 As I mentioned in an earlier chapter, people consider articles containing mathematics more reliable, even if they only spout nonsense.

133 For an example, see http://bibliotecadigital.fgv.br/dspace/bitstream/handle/10438/12441/TESE%20 -%20ARTHUR%20SCHULER%20DA%20IGREJA%20-%20ENTREGA%20FINAL.pdf?sequence=1

chometric qualities of the ipsatized HBDI, we still face multiple difficulties: the Clemans formula (-1/(m-1) applies to forced choices between two options, whereas the HBDI forces participants to choose 1 out of 4, 3 out of 2, 4 out of 16, etc. as I explained more extensively in the Belbin section. The intercorrelation matrix of a series of ipsative scales is an artifact that is likely to produce distortions in the scores, in turn making meaningful interpretations impossible:

“If researchers carry out a factor analysis on such artificial matrices, they will find even more peculiar effects, as have been well described by other authors (Baron, 1996; Bartram, 1996; Clemans, 1956; Dunlap & Cornwall, 1994; Hammond & Barrett, 1996; Hicks, 1970; Johnson, Wood, & Blinkhorn, 1988; Tenopyr, 1988). These effects are of a statistically technical nature (such as, production of artificial bipolar factors; sometimes commonalities equal to 1 and the absence of specific variance; the lower the scales, the greater the artificial distortion; and the like).” (this is my own text from the Belbin chapter)

In sum, the forced choice format is really about ‘forcing one to choose under restriction of possible answers and scores, often euphemistically called ‘a preference.’ At best, you can use an ipsative scoring format for ‘within person’ comparisons, not for ‘between person comparisons’:

● With a ‘normative’ score, people are compared to a population norm,134 thus offering an objective comparison.

● Ipsative scores can’t to be related to a population norm: there is no objective comparison to any other individual. They can only reveal ‘a within a person’ choice if forced to choose.

If this wasn’t difficult enough, Bunderson et al. used 135 ipsative scores as a “ continuous score scale” 136 and then further “standardized” these (p. 15) by transforming them into zscores (p. 20). A psychometrician I consulted advised me to explain the difference between nominal, ordinal, interval, and ratio variables and the acceptable mathematical calculations for them. However, I have decided not to, as it is already difficult enough and you can find adequate explanations on the internet. I will try to give you the bottom line, but I have to admit it is a bit technical (if you give up, just read the last sentence of the following paragraph).

Psychometric experts agree that even a normative score like a 5-point Likert scale is an ordinal score and not a continuous score137—which has fixed, and often narrow, intervals. If you add all the scores of the items belonging to one scale, it is acceptable to treat the total score obtained from a 5 or 7-point Likert scale as continuous scores. It is argued that the underlying theoretical construct is an interval measure that cannot be measured directly or with only one item (e.g. you can measure temperature directly, but you cannot measure the psychological construct of extraversion directly). So, if you have enough items in a Likert scale, you can treat their total score as a continuous score. Ipsative scores, however, show no intervals or continuity at all and are ordinal scores, expressing a simple order or rank. Whereas it is acceptable to make calculations on normative scale score totals, this is

134 This is a ‘between person comparison.’ Often one’s position is shown on a bell or Gauss curve.

135 They use the word “computed,” which is an inappropriate use of this word. They used a technique that should not have been used on ipsative scores.

136 You need quite a narrow and fixed interval for continuous scores. Ipsative scoring results in an ordering of the items, e.g. item B is more important than A Whereas in a 5-point normative Likert score, you would for example have answered A= 4 and B= 5. Ipsative scores do not reflect subtle differences.

137 Mathematically speaking, a continuous score has an infinite number of possible values between two points (as you can express numbers in 1, 2, 3, 4, 5, etc. decimal places).

not the case for ipsative scale score totals. An example will help make this crystal clear. I think everyone would agree that a measure of 100 equally spaced interval variables (like the temperature from 0–100°C) follows a continuum (thus a ‘continuous score’). A scale consisting of 15 questions using a 5-point scale can offer a nuance of 60 points (i.e. 75–15) to measure the underlying psychological construct, so we can treat them as almost continuous scores. This is an allowed mathematical calculation. However, this is not the case for a forced choice format. In that case, the scale score has only 15 points and only expresses a rank or order (as in: I prefer apples over pears—but it doesn’t say anything about how much more I like apples). The interval between their values is not constant or continuous. In short, Bunderson et al. used mathematical calculations (e.g. interval techniques, standardization) where they shouldn’t have because their scores were ipsative.

Another peculiar aspect: two scales of the questionnaire were left out (‘Learning and career Choices’ and ‘Hobbies’) as well as some items that did not fit the criterion of ‘preference choice under limitations.’ As demonstrated by Sean Hammond and Paul Barrett in a paper for the British Psychological Society in 1996, dropping some scales “will dramatically affect the interpretation of the factor solution ” (p. 142) . Their research replicated the problems found in other studies using different (ipsative) tests. The decision to leave out certain scales and items influences the factor solution.

By now you can understand that I almost went mad trying to make sense out of the paper. However, I was given reassurance when the psychometric experts I consulted also quickly gave up. They told me it is impossible to understand what the figures mean. One expert summarized the methods in the Bunderson et al. 2014 paper as follows: “Ipsative (scores) are used as continuous (interval) and are then standardized. No one can understand what such figures ultimately mean.”138

The choice of certain techniques is… well, unusual. Forgive me for repeating some sections of text that I wrote for the Belbin section, but the text below is a very important reminder: Dunlap and Cornwall (1994, p. 123) concluded:

“We are left with recommending against the use of principal component, principal factoring, or maximum likelihood factor analysis with ipsative measures. The separation of artifactual bipolar factors induced by ipsativity from any true underlying relationship will be difficult at best, and not worth the danger of a largely incorrect interpretation.” (Bold emphasis my own)

Tenopyr (1988, p. 750) stated:

“And so, it is clear that the internal consistency reliabilities of forced-choice scales are not only interdependent but may also augment each other enough to give the scale user a false sense of confidence in construct interpretation,” and “… a subject for major concern in view of the widespread use of forced-choice inventories in vocational guidance, diagnosis, selection, and research.”

Bearing in mind this advice about ipsative scores, I was surprised by the decision to use a principal component analysis (one form of factor analysis), so I consulted a psychometric expert who pointed out a fact that I had overlooked. Indeed, the decision to use a two-factor solution is not well explained. The paper mentions that these two factors jointly explain 67.7% of the variance. The authors fail to mention how much of the variance the two factors explain separately. Maybe one factor explains most of the variance? Or are they equally

138 E-mail from Danny Rouckhout (University of Antwerp) on November 4, 2015.

strong? We simply don’t know. Most psychometricians now consider Structural Equational Models (SEM) to be a better method for determining how well the data match the theoretical construct (goodness-of-fit), but you must use a homogenous set of data (e.g. one language, one country). From the Bunderson et al. 2014 paper, it becomes obvious that such datasets do exist. Although they acknowledge the HBDI structure should be analyzed using “specific national and linguistic subgroups in order to assess the extent to which translated questions and section formats perform the same measurement functions across translations” (p. 13), the authors have still chosen to “investigate the structure of the HBDI on an international population as a whole. ” (p. 13) Moreover, the authors also acknowledge that “ it is recommended that confirmatory factor analysis should be used starting with homogeneous samples consisting of only one language version” (p. 15), but instead of respecting this guideline, they conduct a CFA on an international sample, consisting of U.S. English, British English, other “international English samples,” French, German, and other languages from Asia and the Middle East. So, even though the data are available and SEM is a perfectly good method, they still used factor analysis, despite recommendations against it. That leaves me wondering and guessing as to why they did this. Did they absolutely want to prove something?

It also strikes me that Bunderson and his colleagues tried to test the underlying structure of four scales (based on ipsatized scores), when most psychometric experts strongly advise against this practice. Suffice it to say, the following quoted conclusion from Helen Baron (1996, p. 55) says it all:

“The artificial interdependence between ipsative scores does effect their psychometric properties. These effects tend to be greatest where an instrument has few scales and there are higher underlying intercorrelations between the constructs measured. However, with around 30 scales, ipsative measurement does provide some interpretable psychometric parameters. Much larger numbers of scales seem to be needed before factor analytic results resemble those provided by normative data.”

In the case of the HBDI, only 4 scales (containing ipsative scores) are used, so the underlying structure cannot be adequately tested, according to the vast majority of psychometric experts.

Since Bunderson is also obviously a believer of learning styles, I cannot help but think of the old adage ‘birds of a feather flock together.’ It seems it was quite convenient for the HBDI to ask Bunderson to conduct a statistical analysis for a second time. This leads to the obvious and simple conclusion: the selected statistical methods and reporting are far from convincing. I still am patiently waiting for independent research on the statistical features of the HBDI questionnaire. Or shouldn’t we spend tax money on an entirely false and impossible theory about the brain and thinking styles? Right.

To put it simply (but correctly)

Of course, all that being said, it is extremely difficult to construct a questionnaire to measure four quadrants in the brain or dominant thinking styles that simply don’t exist. For the

layperson or people whose statistics are a bit rusty, I checked with Danny Rouckhout, a psychometric expert from the University of Antwerp, whether a simplified story could be told. Here it is, written from my hypothetical perspective:

● I have an idea that there are four thinking styles. I hypothesize they are analytic, organizing, communicating, and creative.

● I write a number of items which I think they will reflect one particular thinking style (meaning I write these items with assumptions in the back of my mind).

● I create a questionnaire, but I force people to choose between options (only 25% of the items may be chosen—I offer these items in series of 4, 12, 16 and 20).

● This gives me the results I had hoped to obtain: two first-order factors and two secondorder factors. I ignore the fact that factor analysis on ipsative items is an unacceptable computation. I use ipsative items as continuous items.

● I consider these results to be evidence for four thinking styles.

● My circular reasoning has delivered the desired result—my only problem is… it is circular reasoning. The dog chases its own tail.

● But I take an even bolder step: I name the factor left-right (brain hemispheres) and upper-lower (limbic). I tell my audience that the four factors or quadrants I found offer evidence for brain dominance. I do all this without conducting (expensive) brain research of course.

● It gives me great satisfaction that my model and questionnaire are used for over 30 years without any major opposition. The naive and gullible public is lazy, so I am able to get away with it.

The empirical score: -3. If we were to look at only the psychometric properties of the questionnaires, we would have to conclude it has poor psychometric properties (because of its ipsative nature, unacceptable computations, etc.), thus resulting in a score of -2. But, since the theory of left versus right brain dominance has been refuted by the vast body of empirical findings, it ultimately deserves a score of -3.

Why do people believe HBDI can ofer them valuable insights?

Neuroscientists think the left-brain versus right-brain dominance myth dates back to at least the 1800s. The myth was later fueled by research conducted in 1960 by Roger Sperry and Michael Gazzaniga with regards to epilepsy treatment, in which they cut the corpus callosum, a band of neural fibers that connects the two hemispheres. Michael Gazzaniga himself said about this left versus right brain myth in an interview with Kayt Sukel: “It took off and really had its own life,” he chuckled. “And it makes sense if you think about it in terms of a very easy way to explain what you knew about brain mechanisms and cognitive abilities. But it’s overly simplified and overstated.”

“There are some kids who visualize problems and other kids who verbalize them. And some educators use those terms, visualizer and verbalizer. That reality has been mapped on the right brain/left brain anatomy as an explanation. But that’s where it falls down. Because the actual neural mechanisms for how these cognitive strategies work are much more complex than that. Cognition, in general, is much more complex than that. That’s what we’ve learned over the years and continue to learn as we study hemispheric differences. It’s all just a lot more complicated than we ever thought.”139

Ned Herrmann and his institute always referred to Roger Sperry’s work, probably in an effort to give the theory a luster of respected science . They claimed that Sperry’s Nobel Prize-winning research ‘validated’ the model or that he was a ‘key contributor.’ This is a clever way to suggest to the public that Roger Sperry made direct contributions. A blatant lie would have been to assert that ‘Roger Sperry validated our whole brain theory/our brain dominance theory.’ However, this subtle, suggestive wording is likely to have helped them establish what became a flourishing business. Another of their suggestions was that Victor Bunderson did some serious work. I am not downplaying his possible skill in statistics, but he is neither a medical doctor nor a neuroscientist or a brain researcher. The first report he

139 Kayt Sukel in “When the Myth is the Message: Neuromyths and Education” (Dana.org).

produced is, at the very least, besides the question and extremely chaotic, while his methodological choices were questionable. To top it all off, the HBDI website claims that some 60 doctoral dissertations, or more, have been conducted on or using the HBDI.

To the unsuspecting reader, all these authority arguments may sound very impressive. As most HR people and participants are not brain specialists, neuroscientists, or psychologists, they may not feel the need to dig into the research papers or try to contact some of the ‘contributors’ as I did. The strategy of using authority arguments increases credibility at face value. As I outlined in the introductory chapter, we are all highly vulnerable to authority arguments, whether real or false. Fact-checking everything would make life almost impossible (is he a real dentist, is she a real doctor, is that really a lawyer who studied at law school…?). But the fact that we often rely on true or false authorities makes us vulnerable to fraud and misleading statements or mistakes.

Another classical marketing trick is to create nice product packaging. Circles, quadrants, and colors are extremely attractive to both consultants and their customers. Meanwhile, other often used persuasion techniques include the Bandwagon effect (social proof):

“Our Whole Brain® Thinking solutions are trusted by the world’s most admired companies.”

“With over 2 million Thinkers globally, our solutions are used by 97% of Fortune 100 companies. We also have 16 offices worldwide and the HBDI® is available in 20+ languages so you can align your culture.” (HBDI website)

And, like so many other instruments, any result given to participants is likely to create the Forer effect—an erroneous acceptance of positive generalizations because one thinks it reflects the particulars of their personality (see Part I). Think back to Forer’s conclusions from his experiments: “A minimum degree of correspondence between the sketch and selfevaluation appears to engender an attitude of acceptance of the total sketch and this attitude of acceptance is carried uncritically to the test instrument” (1949, p. 123) This mix of psychological mechanisms, vulnerabilities, and bias offer a good explanation for the relative popularity of HBDI.

How likely is it this theory will ever prove to be valid?

Using the current scientific mapping techniques for the brain and its functional networks, the idea of the brain being simply divided into four neat quadrants has been totally refuted and will never be proven valid. One hemisphere is not generally dominant over the other. Rather, each hemisphere of the brain dominates the other hemisphere to a certain extent for certain functions. Language, math, and logic typically involve the left hemisphere more, whereas spatial abilities, facial recognition, music, and visual imagery involve the right hemisphere.

We can deduce from his later writings that Ned Herrmann eventually acknowledged this. He probably wanted everyone to consider the “whole brain model” or “brain dominance” in a purely metaphoric sense. But neither he, nor his institution, ever changed the name of the model or the questionnaire.

■ Original sources consulted

Bunderson, V. (1989). The validity of the Herrmann Brain Dominance Instrument. Unpublished and non peerreviewed manuscript. Summary to be found on the websites of the Herrmann Institute.

Bunderson, V., Olsen, J.B., & Wendt, D.C. (2014) Validity of the HBDI Four-Quadrant and Left-Right Structure: a Structural Validation Study Using New Millennium Database. Unpublished and non peer-reviewed manuscript. found on the website of the Herrmann Institute. All authors are employed at the EduMetrics Institute founded by Victor Bunderson.

Clayton, P., & Kimbrell, J. (2007). Thinking preferences as diagnostic and learning tools for managerial styles and predictors of auditor success. Managerial Finance, 33(12), 921–934.

Clemans, W. V. (1956). An analytical and empirical examination of some properties of ipsative measures. Psychometric Monographs, 14. Richmond, VA: Psychometric Society.

Deshpande, P., & Baxi, S. (2011). ST-IM model: a whole brain approach to implementation. Industrial and Commercial Training, 43(4), 228–238.

Hammond, S., & Barrett, P. (1996). The psychometric and practical implications of the use of ipsative, forcedchoice format, questionnaires. In: Book of proceedings (occupational psychology conference of the British Psychological Society, 135–144.

Meade, A. W. (2004). Psychometric problems and issues involved with creating and using ipsative measures for selection. Journal of Occupational and Organizational Psychology, 77, 531–552.

Moeller, J. (2015). A word on standardization in longitudinal studies: don’t. Frontiers in Psychology. 6, article 1389.

Nielsen, J. A., Zielinski, B. A., Ferguson, M. A., Lainhart, J. E., & Anderson, J. S. (2013). An evaluation of the leftbrain vs. right-brain hypothesis with resting state functional connectivity magnetic resonance imaging. PloS one, 8(8), e71275.

Rowe, F. A., Waters, M. L., Thompson, M. P., & Hanson, K. (1992). Can personality-type instruments profile majors in management programs? Journal of Education for Business, 68(1), 10–14.

van Oordt, M.L., Van Oordt, T., & du Toit, P. (2014). Are two teachers better than one? meditari Accountancy Research, 22(2), 166–185.

The Skeptic’s Dictionary Newsletter, April 6, 2008 (see www.skepdic.com).

APA PsycNET database.

EBSCO Business Source Elite database.

ProQuest ABI/INFORM database.

Turn static files into dynamic content formats.

Create a flipbook
Myth 11 - Herrmann Brain Dominance Instrument by Uitgeverij Lannoo - Issuu