Iranian EFL Learners’ Online Self-regulated Learning, Use of Communication Strategies, Test Anxiety and Online Speaking Test Performance: A Structural Equation Modeling Approach

The obligatory prevalence of online education during the COVID-19 pandemic has drawn researchers’ attention to the challenges involved in foreign language pedagogy in such virtual educational contexts. Against this backdrop, this study investigated the impact of online self-regulated learning, use of communication strategies, and test anxiety on Iranian English as a Foreign Language (EFL) learners’ online speaking test performance. For this purpose, 132 EFL learners were given the e-Oxford Quick Placement Test and the speaking part of a sample A2 Key and B1 Preliminary test. Next, translated versions of the given measures were administered to the pre- and intermediate EFL learners and the obtained data were subjected to Structural Equation Modeling analyses that verified strong links between online self-regulated learning and the use of communication strategies, test anxiety and online self-regulated learning, and test anxiety and EFL learners’ use of communication strategies. Furthermore, the direct impacts of online self-regulated learning and use of communication strategies on learners’ online speaking test performance were verified; however, test anxiety was found to indirectly impact the learners’ online speaking test performance through its negative effect on EFL learners’ online self-regulated learning and use of communication strategies. In addition, online self-regulated learning turned out to be the strongest predictor of the learners’ online speaking test performance. As for the implications of the findings, it appeared that the attested model lends support to Bachman and Palmer’s (1996) language use framework illustrating test performance as a vulnerable construct affected by test takers' attributes and features of the test tasks and the impact of construct-irrelevant factors like test takers’ personal characteristics on their test performance.