The project report summarized here describes the interrelationships of the individually administered tests given to 820 Head Start children from three urban sites in the second year of a longitudinal study. These results are compared with those obtained in Year 1, the year prior to the students' attendance in a group program. The tests tapped cognitive skills (e.g., general knowledge, verbal and perceptual-motor abilities, reasoning skills, Piagetian-based conservation development), cognitive styles (e.g., analytic abilities, reflection-impulsivity), and a variety of personal and social behaviors (e.g., risk-taking, cooperation, self-concept, and play preference). In both years, analyses of the test data yielded two main dimensions: (1) a general ability dimension, and (2) a response tempo dimension. Some task-specific styles and behaviors were also defined. Year 2 analyses also revealed a spontaneous verbalization factor unrelated to the major verbal competency measure. In addition, there was some evidence in Year 2 for generalization of specific personal and social behaviors (compliance and the desire to please) across tasks. Both economic eligibility and later selection for Head Start were associated with cognitive performance differences in Year 1, and significant differences remained in Year 2. High similarity in the factor structure across years was shown. (Author/KM)