Multidimensional ability tests and culturally and linguistically diverse students: Evidence of measurement invariance

Ability tests are used by teachers to provide additional context for interpreting student achievement and as a tool for differentiating instruction to the cognitive strengths and weaknesses of students. Tests that provide the most useful information for these purposes measure school-related content domains including verbal and quantitative reasoning. However, there is concern that verbal content affects validity for culturally and linguistically diverse students. In this study, the structure of a multidimensional ability test of verbal, quantitative, and nonverbal reasoning skills was explored in three groups of students who varied in language and cultural background. Configural invariance and invariant factor loadings were supported, but the verbal battery showed weaker relationships to the other batteries and reduced variability for English learners. Results indicate that battery-level scores are appropriate for all students, but that accounting for educational opportunity may be required for interpreting scores.

► The Cognitive Abilities Test (CogAT) was administered to ELL and non-ELL students.
► Configural invariance for factor structure confirmed for ELL and non-ELL students.
► Verbal factor had smaller variance and weaker factor correlations for ELL students.
► Results support use of battery-level scores for ELL and non-ELL students.
► Interpreting scores may require use of local norms based on opportunity to learn.