RESUMO
193 basically healthy working-class and middle-class mothers and their infants participated in a 4-year longitudinal study which focused on the relative potency of several clusters of variables for predictions of intellectual and language outcome during the preschool years. The major results were: (1) Measures of perinatal or infant physical status were extremely weak predictors of 4-year IQ or language. (2) Assessments of child performance were poor predictors prior to 24 months, but excellent predictors from 24 months on. (3) Assessments of mother-infant interaction and general environmental quality were among the best predictors at each age tested, and were as good as measures of child performance at 24 and 36 months in predicting IQ and language. (4) Measures of the family ecology (level of stress, social support, maternal education) and parent perception of the child, especially when assessed at birth, were strongly related to child IQ and language within a low-education subsample, but not among mothers with more than high school education. Patterns of prediction were similar for 48-month IQ and 36-month receptive language; predictions were notably weaker for 36-month expressive language.
