In order to test the hypothesis that middle-class subjects will spend more time on a difficult task than will working class subjects, 40 adult white married females were randomly selected from two small towns in north-central Oklahoma and given independently and in random order a logico-manipulative task and a motor-manipulative task. (Author/JM)

Examines the possibility that specification errors contributed to the Schuman et al. (1985) findings of a weak relationship between study time and college grades. Analyzes main and interactive effects, measures of quantity and quality of study, and various context-specific models of college grades. Results challenge previous findings. (Author/SV)