Evaluating Invariance in Test Performance for Adolescent Learners of English as a Foreign Language

Abstract

In this study, we assessed the invariance in the factor structure underlying English‐language proficiency for two groups of adolescent learners in Japan: students in middle school (ages 13–15 years) and students in high school (ages 16–18 years). Language proficiency was measured using the TOEFL Junior® Comprehensive test, an assessment designed to measure the English skills of adolescent English learners in non‐English‐speaking countries. The study results indicate that a correlated 4‐factor model corresponding to the 4 language abilities of reading, listening, speaking, and writing best represents the nature of language proficiency in the 2 groups of adolescent English learners. Moreover, the factor structure is invariant across these 2 groups. However, there is a statistically significant difference in performance on the reading construct. The results are consistent across 2 random samples, thus providing confirmatory evidence of model invariance. This study provides empirical support for the current score‐reporting practices for the TOEFL Junior Comprehensive test and suggests that the test scores have the same meaning across test takers from middle school and high school.