Measurement invariance of Bildiren non-verbal cognitive ability (BNV) test across gender, grade level, age, and ethnicity

Abstract

Testing measurement invariance is crucial when working with diverse samples and making critical decisions about individuals, as accurate and meaningful cross-group comparisons are essential. This study investigates the measurement invariance of the Bildiren Non-Verbal Cognitive Ability (BNV) Test across gender, grade level, age, and ethnicity. The analyses utilized data from two samples comprising students aged 4 to 13: Sample 1 (N = 7745), consisting solely of Turkish students, was used for gender (boys and girls), grade level (primary and elementary), and age (4–5, 6–7, 8–9, 10–11, and 12–13 years old) comparisons, while Sample 2 (N = 1719), comprising both Turkish and Syrian students, was used for ethnicity comparisons. Measurement invariance was assessed separately for each group using a multigroup confirmatory factor analysis (MG-CFA) approach. The results indicated scalar invariance, demonstrating that factor loadings and thresholds were equivalent across gender, grade level, age, and ethnic subgroups. These findings support the comparability of factor means across these groups. The results provide validity evidence for the BNV Test scores, supporting valid comparisons across subgroups and enabling their use for diagnostic, evaluative, and selection purposes. This study underscores the importance of assessing measurement invariance to ensure test scores are interpreted fairly and accurately across diverse test-taker groups.