Simultaneous Screening of Child Health and Development: Further Test Development, Field Testing, and Predictive Validity

摘要

The project was designed to develop and refine preschool pediatric screening tests. Standard psychometric procedures for the development and evaluation of tests were employed. These included assessments of item sensitivity, reliability (test-retest, inter-observer, and internal consistency), concurrent validity, and predictive validity. Both correlational and crossclassificatory techniques were employed to assess the validity of the tests. The data indicate that START is a good predictor of poor school outcome at the end of kindergarten, first, and second grade. prediction decreases somewhat across years, however, as might be expected with the increasing time delay between screening and follow-up. Overall, though, more than half of children with poor school outcomes in kindergarten failed START in the spring prior to school entry. Because poor school outcome is a multivariate outcome, the authors attempted to add information from parent questionnaires and other screening test to START to improve prediction. The results of these efforts indicate that combining information from several sources did improve our ability to predict school outcome from prekindergarten screening.