The Handbook of Research Synthesis

Appendix A: Data Sets

APPENDIX A Data Sets Three data sets have been used repeatedly throughout this handbook. TIlls appendix provides brief descriptions of the data sets, references for the research studies that reported the primary data analyses, and integrative research reviews that examined the studies. 1. DATA SET I: Studies of Gender Differences in Conformity Using the Fictitious Norm Group Paradigm Eagly and Carli (1981) conducted a comprehensive review of experimental studies of gender differences in conformity. The studies in this data set are a portion of the studies that Eagly and Carli examined, the studies that they called "other conformity studies." Because all of these studies use an experimental paradigm involving a nonexistent norm group, we call them "fictitious norm group" studies. These studies measure conformity by examining the effect of knowledge of other people's responses on an individual's response. Typically, an experimental subject is presented with an opportunity to respond to a question of opinion. Before responding, the individual is shown some "data" on the responses of other individuals . The "data" are manipulated by the experimenters , and the "other individuals" are the fictitious norm group. For example, the subject might be asked for an opinion on a work of art and told that 75 percent of Harvard undergraduates liked the work "a great deal." Several methodological characteristics of studies that might mediate the gender difference in conformity were examined by Eagly and Carli (1981) and by Becker (1986). Two of them are reported here: the percentage of male authors and the number of items on the instrument used to measure conformity. Eagly and Carli reasoned that male experimenters might engage in sextyped communication that transmitted a subtle message to female subjects to conform. Consequently, studies with a large percentage of male experimenters would obtain a higher level of conformity from female subjects. Becker reasoned that the number of items on the measure of conformity might be related to effect size for two reasons: Because each item is a stimulus to conform, instruments with a larger number of items produce a greater press to conform; instruments with a larger number of items would also tend to 'be more reliable, and hence they would be subject to less attenuation due to measurement unreliability. A summary of the ten studies and the effect size estimates calculated by Eagly and Carli, and in some cases recalculated by Becker, is given in Table A.I. The effect size is the standardized difference between the mean for males and that for females, and a positive value implies that females are more conforming than males. The total sample size (the sum of the male and female sample sizes), the percentage of male authors, and the number of items on the outcome measure are also given. 2. DATA SET II: Studies of the Effects of Teacher Expectancy on Pup" IQ Raudenbush (1984) reviewed randomized experiments on the effects of teacher expectancy on pupil IQ (see 543 544 APPENDIX A TableA.1 Data Set I: Studies of Gender Differences in Confonnity Total Effect Male Sample Size Authors Items Study Size Estimate (%) (n) King 1959 254 .35 100 38 Wyer 1966 80 .37 100 5 Wyer 1967 125 -.06 100 5 Sampson & Hancock 1967 191 -.30 50 2 Sistrunk 1971 64 .69 100 30 Sistrunk & McDavid 1967 [1] 90 .40 100 45 Sistrunk & McDavid 1967 [4] 60 .47 100 45 Sistrunk 1972 20 .81 100 45 Feldman-Summers et al. 1977 [1] 141 -.33 25 2 Feldman-Summers et al. 1977 [2] 119 .07 25 2 NOTE: These data are from Eagly and Carli (1981) and Becker (1986). also Raudenbush & Bryk 1985). The experiments usually involve the researcher administering a test to a sample of students. A randomly selec.ted portion of the students (the treatment group) are identified to their teachers as "likely to experience substantial intellectual growth. " All students are tested again, and the standardized difference between the mean treatment group and that of the other students is the effect size. The theory underlying the treatment is that the identification of students as showing promise of exceptional growth changes teacher expectations about those students, which may in turn effect outcomes such as intelligence (possibly through intermediate variables). Raudenbush (1984) proposed that an important characteristic of studies that Table A.2 Data Set II: Studies of the Effects of Teacher Expectancy on Pupil IQ Study Rosenthal et al. 1974 Conn et al. 1968 Jose & Cody 1971 Pellegrini & Hicks 1972 [1] Pellegrini & Hicks 1972 [2] Evans & Rosenthal 1968 Fielder et al. 1971 Claiborn...