A sensor-lite anxiety detector for foreign language learning

Abstract

Foreign language learners face cognitive and emotional interference which reduces their learning acquisition. Foreign language anxiety (FLA) is one of the impediments to learning a new language. It hinders cognitive processes, interferes with attention, and reduces working memory capacity. Detecting foreign language anxiety is the first step toward reducing and eventually overcoming it. Here we present a novel design to detect foreign language anxiety in the context of an e-learning system using sensor-free and sensor-lite metrics. By adopting the Foreign Language Classroom Anxiety Scale (FLCAS) as a pre-test along with language difficulty self-report, system difficulty self-report, and score, we are predicting the current anxiety level of the learner when learning English as a second language using an e-learning system. We found that some components of the FLCAS can effectively predict FLA for listening and speaking exercises. Additionally, FLCAS plus the self-reports and the score are effective in detecting the learner's current anxiety level for all exercise types.