Abstract
This study attempts to empirically examine the predictive power of learner anxiety and motivation and their causality in explaining proficiency in English as a foreign language (EFL) and other interrelated affective variables, including attitudes, self-confidence, and grit via a treatment-based causal model. To achieve this, the study carried out a 12-week experimental intervention employed by 8 EFL teachers with 282 EFL learners divided into four groups: in group 1, learners were exposed to anxiety-regulating strategies, those in group 2 were exposed to motivational strategies, those in group 3 were exposed to both anxiety-controlling and motivational strategies, and those in group 4 were not exposed to specific anxiety or motivation strategies. Data was gathered using questionnaires, classroom observations, and proficiency tests. ANOVA and ANCOVA tests were employed to assess the treatment effects, and confirmatory factor analysis (CFA), structural equation modelling (SEM) and partial least squares structural equation modelling (PLS-SEM) analyses were deployed to test a causal model for the relationship between affective variables and second language (L2) proficiency in the group in which the experiment was most effective. Significant direct paths were captured from all affective variables to L2 proficiency. Motivation and anxiety exerted significant direct, indirect, and high total effects on L2 proficiency, confirming them as the best predictors and the causal variables of learners' proficiency and other affective variables.