Another Look at the Construct Validity of the Gifted Rating Scales: Preschool/Kindergarten and School Forms

This study further examined the Gifted Rating Scales (GRS) at the internal and external stages of test validation by (a) testing structural fidelity as indicated by the number of factors supported as well as the correspondence of the latent factor structure with the scoring model proposed by the authors and (b) examining the external relations of factors underlying GRS ratings with the general factor from two widely used standardized tests of intelligence using the multimethod confirmatory factor analysis (CFA). Results of our analyses revealed the presence of a large general factor on both the GRS-Preschool (GRS-P) and the GRS-School (GRS-S) forms. This factor accounted for 80% of the variance on the GRS-P and 72% of the variance on the GRS-S, revealing that the vast majority of what the GRS measures is the general factor. In addition, coefficient ω for both the GRS-P and the GRS-S was .95 and ωH for the GRS-S was .92, demonstrating that this general factor is measured with a high degree of reliability. Results of multimethod CFAs indicated the general factor underlying the GRS correlated substantially with the general factor underlying the Wechsler scales, which may suggest that the GRS primarily reflects general cognitive ability and not multiple dimensions of giftedness. Implications of these results and directions for further research are discussed