Date of Conferral
A review of research concerning conformity behaviors revealed a need for large sample, multivariate approaches. To that end, seventeen variables were utilized (birth order, chronological agev sex, race, IQ, socioeconomic status, geographic area uf residencer religion, anxiety, dogmatism, rigidity, acquiescence, ego strength, confidence, extroversion/ introversion, impulsivity, and neuroticism) to describe the behaviors of 104 subjects in a Tuddenham-type setting wherein the subjects were randomly assigned to either "Asch-type" (group pressure) or "Crutchfield-type" (expert pressure) treatment conditions. Utilizing a stepwise multiple discriminant function analysis, discriminant weights were derived. The variable of rigidity was not found to be a valid predictor. With the discriminant weights and the sixteen predictive second-order variables, 218 subjects were randomly assigned to treatment conditions end their "conforming" or "not-conforming" behaviors were predicted. The data of eleven subjects who correctly perceived the contrived nature of the exper'iment were· not included. Of the 123 subjects subjected to " expert pressure," 85 were not-conformers. Chi-Square analysis of the "expert pressure" data yields c significant difference at the . 20 Level of Confidence. Of the 95 subjects subjected to "group pressure," 45 were conformers and 50 were not-conformers. Chi-Square was not significant.. Conformers to "expert pressure" were correctly identified 74 per cent of the time; conformers to "grcup pressure" were correctly identified 62 per cent of the time. Thus, it appears that conforming behaviors are predictable providing a multivariate method is used conjointly with a discriminant function analysis of the data. The most difficult behavior to correctly predict is not-conforming to "expert pressure, 11 with but 32 per cent correct. Too, but 40 per cent of the not-conforming to "group pressure 11 was correct! y p;edi cted. Conformers to 11expert pressure" were seen as typically being first-born, 20.8 year old Caucasian males or females with an IQ of 114. Their parents were high school graduates or above, and were either professional persons, business proprietors, or managers, etcetera. Their geographic area of residence was Rural East North Central United States, but this finding may be due to sampling error. Protestants were more frequently observed (n = 54) than Roman Catholics (n = 28). The conformers were slightly above average in anxiety and dogmatism. Average acquiescence 1 ego strength, and confidence were noted. low impulsivity was seen, and they were slightly extroverted. As a group, they were more neurotic than an average subject. Not-con formers to "expert pressure" were seen as first-born 21.5 year old Caucasian maler. with an IQ of 119. No clear socioeconomic status was noted due to extreme scatter. The geographic area of residence was Rural East North Central United States. They were predominantly Prctestant subjects. They were less anxious than conforn.ers, less dogmatic, and less acquiescent. They did not differ significantly in ego strength, confidence or impulsivity. However, they were more introverted and decidedly less neurotic than their conformer counterpart. Conformers to "group pressure" were equally likely to be either a first or second-born 21 year old male or female Caucasian Protestant with an IQ of 116. They were more anxious than an average subject. Dogmatism, acquiescence, ego strength, neuroticism, and extroversion/introversion were average. As a group, they were slightly more confident. Subjects who were not-conformers to "group pressure" were seen as secondborn 21.5 year old females with an IQ of 116. Two-thirds of these subjects were Protestant; one-third were Roman Catholic. No obvious socioeconomic status could be ascertained. They were less anxious than conformers to "group pressure." No significant differences were noted in dogmatism, acquiescence, ego strength, confidence, impulsivity or extroversion/introversion. They were less neurotic than their conformer counterpart.