Explains that from a prediction standpoint the confluence model is not very efficient. Very modest increments in accuracy are associated with family configuration variables once chronological age is covaried. Suggests that the major postulates of the theory be tested directly, within individuals and with longitudinal data. (Author/AS)

Continues a previous five-year follow-up of preterm and full-term children by studying the continuity in their intellectual and emotional development. Prematurity was predictive for school adjustment at ages six and seven only when regression was performed on the preterm group, but failed to be predictive when mixed groups of preterm and full-term…