Abstract
The construct validity of the Open Middle Test, a measure of children's social problem-solving ability, was investigated by assessing its relationship to self-control (Self-control Rating Scale) and academic achievement (Wide-Range Achievement Test). Subjects were 113 minority fourth graders from two urban parochial schools. A significant correlation was obtained between Open Middle Test effectiveness scores of first responses and WRAT Reading scores. No other correlations were significant. However, mean differences were significant for Open Middle Test effectiveness ratings of first responses and chosen responses, suggesting possible deficits in the evaluation of self-generated problem solutions. Implications of the results for training programs and for the use of the Open Middle Test as an outcome measure are discussed.