Validity study of intelligence subtests for a Brazilian giftedness assessment battery

Abstract

Considering the importance of validity evidence studies when designing psychological instruments, this study intended on verifying the construct validity of reasoning subtests, through convergent evidence and CFA, as part of a battery for giftedness assessment. A Manova was also applied to evaluate sex and grade differences. The sample was composed of 96 students, in which 49 were female, with 6th grade (n=15), 7th grade (n=12), 8th grade (n=49) students and 2nd year (n=20) students from a public school in the state of São Paulo, aged between 10 and 18 (M = 13.4 years, SD = 1.8). The participants answered the Raven Progressive Matrices Test (general scale) and reasoning subtests that belong to the Battery for Giftedness Assessment (BAHA/G). The results from Pearson's correlation indicated convergence as the majority of the factors composed in BAHA/G showed positive and significant correlations with the Raven test factors, and the CFA displayed two latent variables with strong correlations, particularly among their total score (r=.976). With these results, we found the moderate relationship between these instruments. Further studies are recommended on other types of validity evidence of these subtests to confirm its psychometric qualities.