Behavioral Analysis and Identification of Gifted Children

Abstract

Summary Two samples, one gifted, the other “normal,” of 30 children each (11 to 13 years of age) were matched on age, sex, race, and father's occupation. A parent of each child completed the Children's Behavioral Classification Project instrument, and a profile of 30 scaled scores was obtained for each protocol. A discriminant analysis on the basis of the profiles classified 50 of the 60 Ss correctly into their respective sample category, either gifted or normal. Though the results of a multivariate analysis of variance showed no significant differences between the groups when the 30 profile factors were viewed as a whole, univariate analyses on the separate factors yielded significant group differences on several related factors. These were interpreted with caution. Additional factors displayed significant sex differences and one demonstrated a group-by-sex interaction effect. Coefficients of pattern similarity (the rp ) were computed among all pairs of the 60 Ss. An unweighted average linkage cluster analysis grouped 34 of the 60 Ss into six clusters which accounted for 26 of 30 possible gifted children. After an item analysis was performed on the original data, a 42 item subscale of the CBCP questionnaire was constructed which correctly classified 31 of 40 individuals in a cross-validation sample.