Assessing mathematics anxiety with the AMAS: Measurement invariance across four countries

Mathematics Anxiety (MA) affects people of all ages, influencing academic success and daily life. The Abbreviated Math Anxiety Scale (AMAS) is a widely used self-report questionnaire measuring MA and has been psychometrically evaluated in different languages and age groups. However, its cross-cultural measurement invariance, namely whether AMAS items are perceived the same way by people across different countries and cultures, is unknown. Such measurement invariance is a precondition for meaningful cross-cultural comparisons. This study investigated cross-cultural measurement invariance of the AMAS in UK, German, Italian, and Polish university students, using data from 3610 participants from the "Big AMAS database". We found evidence of partial invariance; only one item (item 8) was perceived differently across countries. Partial measurement invariance indicates that latent factor means of the AMAS can be meaningfully compared across the four considered countries. However, a notable methodological consideration arose - the measurement invariance results heavily depended on the estimator used. This study emphasises the importance of testing for measurement invariance across groups to enable meaningful cross-group comparisons, and it highlights the critical role of the estimation method selection and its impact on the conclusions of psychometric studies.