Item Fairness of the Nonverbal Subtests of the Stanford-Binet Intelligence Test, Fifth Edition, in a Latina/o Sample

Abstract

Every widely used psychological assessment instrument is under scrutiny in terms of cultural fairness. The expectation of the reduced-language (Nonverbal) section of the Stanford-Binet Intelligence Scales, Fifth Edition (SB5; Roid, 2003) is that language ought not to be a modifying factor in terms of final score. The purpose of the present study was to explore the effects of acculturation on performance on the Nonverbal subtests of this commonly used standardized measure of intellectual functioning in three groups: Latinas/os living in the U.S. four years or less, Latinas/os living in the U.S. five years or more, and Caucasian/White Non- Hispanics. The study explored whether there was evidence of differential item function (DIF) on SB5 nonverbal subtests for these groups. An analysis of variance was the procedure used for testing the null hypothesis, that the means of the three populations would be equal. It was expected that scores for each of the participant groups would be normally distributed. Group differences that were statistically significant at the 0.01 level were examined for potential unfairness. This study employed archival data from the sample of the Nonverbal subtests of the Standardization edition of the SB5. The stratification variables were age, sex, and race/ethnicity, matched to census percentages. Participants were ages 4-17 years and included 17 Latinas/o living in the U.S. four years or less, 20 Latinas/os living in the U.S. five years or more, and 100 Caucasian/White Non-Hispanics. Out of the 223 analyses of variance, two items were significant at the 0.01 level, and one was significant at the 0.05 level. No significant differences were found on testlet scores, factor scores, or domain scores. These limited findings of DIF favor each group on different items, balancing one another out and thus nullifying the overall bias hypothesis. The results of this study suggest that there is little evidence of item bias on the SB5 Nonverbal scale between children and adolescents who are from a Latin country regardless of time in U.S. and Caucasian/White Non-Hispanic children with comparable ages, genders, and socioeconomic status taken from the normative sample of the Standardization edition of the SB5. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone [phone]. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]