Examining Measurement Invariance Using the MIMIC model : Gender Difference in FLCAS

초록

Foreign language anxiety is rapidly growing concern as different levels of foreign language anxiety inevitably affect the acquisition of the second and/or foreign languages. In addition, foreign language anxiety related research is often concerned with comparison of different degrees of perception across groups (e.g., males and females). Although findings from previous research regarding gender differences in the perception of the foreign language anxiety are meaningful, the inconclusive results of how differently males and females perceive the foreign language anxiety have garnered more attention than before. Hence, this study uses the multiple-indicator/multiple cause (MIMIC) approach to evaluate whether there are items violating measurement invariance caused by different gender on the foreign language classroom anxiety scale (FLCAS) with college students' population. The results show that gender difference might be viewed as a source of measurement bias and that female students have higher degrees of foreign language anxiety than male students across three underlying traits (e.g., communication apprehension, test anxiety, and fear of negative evaluation). This study would be the first one scrutinizing construct validity of FLCAS by disentangling gender difference existed in the underlying traits from group differences arising from the measurement non- equivalence. In addition, results of the current study would be enhanced further if the analyses of non-uniform measurement equivalence with larger sample sizes are performed. Furthermore, it would be interesting to explore how items of FLCAS behave with the respect to item response theory.

참고자료

· 없음