| Abstract: | Data from 104 public school children were analyzed to determine the psychometric and conceptual soundness of a new model-based measure of intellectual functioning: the Universal Nonverbal Intelligence Test (UNIT; Bracken & McCallum, in press). The UNIT was designed to assess the construct of intelligence for at-risk children who are limited English language proficient. The UNIT and subtests from the Woodcock-Johnson-Revised Tests of Cognitive Ability (WJ-R; Woodcock & Johnson, 1989) were administered in counterbalanced order using standardized procedures. Spearman-Brown reliability coefficients provided evidence of the UNIT'S reliability: Global score coefficients ranged from .86 (Reasoning Quotient) to .93 (Symbolic Quotient). Confirmatory factor analyses of the UNIT provided evidence for a two-factor structure. Exploratory factor analyses of the UNIT with the WJ-R yielded support for the UNIT'S measurement of memory and reasoning constructs, although the many subconstructs assessed by the WJ-R and the UNIT preclude a simple interpretation. Correlations among UNIT global scores with age and with composite scores of the WJ-R provide additional evidence of the UNIT'S construct validity. |
