Multi-construct argument-based validation of foreign language emotion scales for young language learners

This psychometric validation study addressed three affective instruments—the short-form Foreign Language Classroom Anxiety Scale (S-FLCAS), the Foreign Language Reading Anxiety Scale (FLRAS), and the short-form Foreign Language Enjoyment Scale (S-FLES)—for use with young language learners (YLLs) in the Japanese elementary EFL context. Adopting an argument-based approach to validation, we collected questionnaire and vocabulary size test data from 567 Grade 5 and 6 students and assembled evidence for factor structure, reliability, generalizability, and relations to other variables. Bifactor-ESEM and CFA models supported the intended factor structures for classroom anxiety, reading anxiety, and enjoyment, with composite reliability coefficients indicating adequate internal consistency. Multigroup analyses contributed evidence supporting the scale generalization (measurement invariance) across grades. Convergent and discriminant validity were supported by correlations indicating that anxiety and enjoyment are related yet empirically distinct constructs, while predictive validity was evidenced by structural models showing that L2 proficiency indexed by vocabulary size negatively predicted classroom and reading anxiety and positively predicted enjoyment. Taken together, the findings suggest that these scales yield interpretable and reasonably stable scores among upper-elementary learners and can be used for within-grade analyses and, with some caution, cross-grade comparisons based on raw scores. Methodologically, this study demonstrates the feasibility and value of multi-construct, argument-based validation work and highlights the need for further cross-linguistic and cross-age scale validation to advance foreign language emotion research.