Are English language aptitude tests biased towards culturally different minority groups? Some Israeli findings

The major aim of this study was to examine the validity and crosscultural gen eralizability of the test bias contention as applied to English language aptitude tests, routinely employed for student selection and placement functions within the Israeli academic scene. The bias analysis was based on the English Language Aptitude Test scores of n = 1778 Jewish and n = 1017 Arab student candidates applying for admissions to a major campus in northern Israel, who were administered the test as part of the scholastic aptitude test battery used for college admissions purposes. The psychometric properties of the English Language Aptitude Test were compared for Jewish and Arab subgroups, respectively, via a variety of internal (e.g. factor loading, reliability, standard error of measurement, etc.) and external (e.g. predictive validity, homogeneity of regression, standard error of estimate, etc.) criteria. The test scores show sig nificant, but slight, intercept bias, tending to overpredict the overall first-year GPA of Arab student candidates. Also, test scores appear to be somewhat less reliable measures for Arab compared to Jewish student candidates. On the whole, however, our data provide only marginal evidence for differential construct or predictive validity of English Language Aptitude Test scores as a function of cultural group membership, thus lending a greater amount of gen erality and crosscultural validity to the findings of much of previous research negating the cultural bias hypothesis.