Giftedness. Can it be predicted in infancy?Clin Pediatr (Phila). 1989 May; 28(5):205-9.CPed
A cohort of 200 children, who were followed from birth and periodically evaluated as part of a longitudinal study of child development, was used to determine the ability of the Bayley Scales of Infant Development (BSID) to predict children who would be considered intellectually gifted at 7.5 years. The cohort was predominantly white (91%), upper middle class (70% in Hollingshead class 1 or 2), and had preschool educational experiences (94%). At outcome, 36 children had WISC-R verbal, performance, or full scale IQs greater than 135 and were called gifted. On average, gifted children walked 0.7 months earlier and spoke two-word sentences 2.2 months earlier than the non-gifted group. As a group, gifted children showed statistically significant, but clinically small, advantages for age of walking, age of speaking two-word sentences, BSID, and Stanford Binet IQ. Of the 36 children who later proved to be gifted, only two had BSID IQs greater than 134. Of the four children who had BSID IQs greater than 134, two maintained their superior function but 2 did not. A discriminant function analysis was used to predict giftedness for individual children based on BSID performance. The resulting classification achieved sensitivity = 0.69 and specificity = 0.39. Although groups of infants who will be gifted have higher BSID scores, the BSID cannot be endorsed as a method of identifying individual infants who will later demonstrate superior cognitive function.