An Analysis of Predictive, Convergent, and Discriminant Validity of the Universal Nonverbal Intelligence Test with Limited English Proficient Mexican-American Elementary Students

Students from diverse cultures present challenges to psychologists and educators. One of those challenges is to validate the current use of test scores with these students. The goal of the present study was to provide information on the validity of using a specific measure of ability with a specific cultural group. The research analyzed correlation coefficients between scores obtained on the Universal Nonverbal Intelligence Test (UNIT; Bracken & McCallum, 1998) and scores obtained on the reading and math portions of the Stanford Achievement Test-10th Edition (SAT-10; The Psychological Corporation, 2003) by Limited English Proficient (LEP) Mexican-American elementary school students. The study used a follow-up with selection by existing system predictive design (Guion & Cranny, 1982) of all elementary school students who met study criteria for a three year period in a small agrarian community. Correlation coefficients between the UNIT Full Scale and area scores and SAT-10 reading and math scores were computed. Standard adjustments for restrictions of range were completed. Coefficients were then examined for significance and effect size. Differences between correlations were also examined for significance and relative strength. The data documented predictive, convergent, and discriminant validity with the targeted population for UNIT Full Scale IQ. Implications and limitations of the research are discussed.

The LEP sample obtained corrected correlations of the UNIT Full Scale IQ and
the WLPB-R Broad Reading, Basic Reading and Reading Comprehension
Clusters of .32, .12 and .39 respectively. The UNIT Symbolic Quotient correlations
with ...