Confirmatory factor analysis of the Reynolds Intellectual Assessment Scales (RIAS) in Canadian children.

The purpose of the current study was to assess the validity of the Reynolds Intellectual Assessment Scale's (RIAS) index and subtest scores by examining whether the RIAS conforms to Carroll's ( 1993 ) three-stratum theory of cognitive ability and the Cattell-Horn Gf-Gc (Horn & Cattell, 1966 ) theory of intelligence upon which it was based. Factor structures of RIAS scores from typically developing (n = 187) and mixed clinical groups (n = 164), 4-18 years old, were compared using confirmatory factor analysis. The results were mixed, with some findings supporting the validity of the Composite Intelligence Index (CIX), Nonverbal Intelligence Index (NIX), and Verbal Intelligence Index (VIX) while others suggested that the CIX and NIX in particular should be interpreted with caution, depending on the population in which the RIAS is being used. The functioning of the two nonverbal subtests in all analyses suggested that the RIAS's nonverbal subtests are most problematic, with the greatest impact on the RIAS's validity as a measure of overall, verbal, and nonverbal intelligence. The RIAS's use as a general cognitive screener is supported but caution is recommended if the RIAS is being used to make important diagnostic, qualification for services, forensic, or placement decisions.