General-Factor and Specific Variance in the WPPSI-R

Abstract

Although theorists disagree on the structure of intelligence, many researchers and clinical professionals have documented the practical importance of psychometric g, a general factor that has appeared consistently in factor analytic studies of cognitive batteries. Given the widespread use of each of the Wechsler scales, an understanding of the proportion of general-factor variance across the preschool, child, and adult versions of the scales is important for both research and clinical assessment. Because the Wechsler Preschool and Primary Scale of Intelligence-Revised (WPPSI-R) is a widely used early-childhood battery and represents a substantial revision of the WPPSI, an updated analysis of subtest variance was needed, especially with the addition of the new Object Assembly subtest. The WPPSI-R standardization sample (N = 1,700) was employed to assess the proportion of general-factor (g), common-factor, specific, and error variance for each subtest and the battery as a whole, using methodology comparable to previous studies. Subtests that had the highest g loadings (Information and Arithmetic) and the lowest loadings (Animal Pegs and Mazes) were identified. The subtests with lowest specificity were Information and Object Assembly. WPPSI-R then was compared to WPPSI and other Wechsler batteries, and a trend toward higher g saturation in adult vs. children's batteries was noted.