Social Insight and Evaluative Reliability

Many theories of social development (2, 4, 5) suggest that individuals who are efficient at recognizing and understanding the attitudes and motivations of others are also effective at making social evaluations. To investigate this relationship 64 students in introductory psychology classes at Austin Peay State University, 20 males and 44 females. completed the Chapin Social Insight Test (1, 3) and also rated the atmosphere in their families while they were growing up on 10 bipolar semantic differential scales. Insight scores ranged from 11 to 32. Subjects were divided at the median of the insight scores (Mdn = 21.0), with the same number of males and females in each group. The measurement errors on the semantic differential scales for these groups were then compared. Although there is no single criterion for discerning effectiveness of social evaluations, at least one useful index is the individual's reliability. Subjects with high insight scores had somewhat lower reliabilities on the average than those with low scores; M = 9.62 and 12.19, respectively. Since the Bartlett test (XI' = 4.70, p < .05) indicated that the variability of measurement errors was significantly greater among subjects with low insight scores (2 = 126.23) than among subjects with high scores (2 = 63.071, a Kruskal-Wallis test was used to analyze the data. The results indicate that the average rank of the errors for subjeces with low insight scores was significantly greater than for subjects with high scores (H = 194.95, p < .001). These findings support the hypothesis that individuals with greater social insight are more reliable judges of social situations and also suggest that the insight scale might measure role-taking among adults.