Abstract
On the basis of Chapin Social Insight Test scores from two samples of male and female undergraduates representing a university population in consecutive years, estimates of split-half (ns = 66, 128), and test-retest (ns = 44, 99) reliability were obtained. In addition, for 63 students in the second sample, correlations between test scores and course grades were calculated. Test-retest reliabilities were .69 and .66, split-half reliabilities were even lower (.20 to .50), and validity coefficients between Chapin scores and grades in behavioral science (.09) and nonbehavioral science course (.06) were statistically nonsignificant. In addition, although they were not restricted in range, the mean scores for both sexes fell about one standard deviation below the relevant norms. These data provide further evidence for the separation of social and academic intelligence, but they also suggest that the Chapin test should be used with caution.
Get full access to this article
View all access options for this article.