1
Solution Manual for Research Design and Statistical Analysis, Fourth Edition By Caren M. Rotello, Jerome L. Myers, Arnold D. Well, Robert F. Lorch, Jr. Chapter 1-25 RDSA4e CHAPTER 1 Solutions to Exercises in R 1.1
1.2
1.3
(a) The study is an experiment. Participants were randomly assigned to each condition. (b) The sample might best be characterized as a random sample of (presumably) overweight volunteers. The volunteer aspect is important because it's not clear that the effects of the diets would be the same if used by individuals less motivated to lose weight. (c) One alternative approach is to observe individuals who have selected a particular diet. This might be done by advertising for individuals who are on the diets of interest, requesting they volunteer as participants in a study and then simply assessing their weight. This not as attractive an option as the experiment for several reasons. First, it’s not an experiment, so conclusions about the causal role of diet in weight are not possible. Second, the researcher wouldn’t have baseline information on weight and health prior to the start of the diet. Some other concerns are that the targeted diets may not be equally-well represented in the volunteer sample, and individuals who chose different diets may differ systematically with respect to other factors that could their weight. (a) It is reasonable to assume that the results hold for children from a similar home and school environment, though before drawing conclusions about the two methods, it would be wise to test generality by replicating the experiment at other similar schools. With respect to the broader population of third graders, we should be still more cautious about generalizing until further research is carried out. Factors such as experience with computers and parental involvement in arithmetic learning may be different for other segments of the population. (b) A major problem in such studies is equating the amount of practice in the two conditions. Another variable is teacher experience and competence. Assigning the same teacher to both classes is a possible solution but not necessarily the best one if the teacher is biased in favor of one of the methods. The bottom line is that considerable effort must be taken both with the materials and the teachers prior to the experiment. (a) There are several factors other than long hours in day care that may increase aggressive behavior. Two possibilities are that (1) parents may be more likely to place more aggressive children in day care to provide a socializing experience or for other reasons; (2) parents who are more stressed may foster aggression in their children and may place those children in day care for longer hours to reduce their stress or to allow them to work more/earn more money. (b) Ideally, a measure of aggressive behavior prior to the first enrollment in day care should be obtained. If variation in aggressive behavior as a function of hours/day in day care manifests itself only after the first placement in day care, there is evidence that