THE CONSTRUCT VALIDITY OF LANGUAGE APTITUDE

Abstract

A meta-analysis was conducted to examine the construct validity of language aptitude by synthesizing the existing research that has been accumulated over the past five decades. The study aimed to provide a thorough understanding of the construct by aggregating the data reported in the primary research on its correlations with other individual difference variables and with second language (L2) achievement. A total of 66 studies were retrieved that contributed effect sizes based on 109 unique samples and 13,035 foreign language learners. The results revealed that (1) aptitude was independent of other cognitive and affective factors: it was distinct from motivation, had a negative correlation with anxiety, and overlapped with, but was distinguishable from, intelligence; (2) executive working memory was more strongly associated with aptitude and aptitude components than phonological short-term memory; (3) aptitude measured using full-length tests was a strong predictor of general L2 proficiency, but it had low predictive validity for vocabulary learning and L2 writing; and (4) different aptitude components demonstrated differential predictive validity for different aspects of learning. The findings are useful for tackling a number of conundrums surrounding language aptitude and shed light on how to reconceptualize the construct and reorient the research.