Gaze Behavior Across Proficiency Levels in Video‐Mediated Second Language Interactional Competence Assessment

Abstract There has been an increasing interest in better understanding the role of embodied resources in second language (L2) assessment contexts and thus expanding the construct of interactional competence (IC). However, multimodal analysis has yet to break new grounds in video‐mediated interactions in digital assessment environments. This study investigates gaze in video‐mediated L2 English paired role‐play assessment interactions collected from 117 randomly paired test takers (Common European Framework of Reference, CEFR levels A2–B1), focusing on how test takers manage gaze behavior when encountering trouble in producing turns (e.g., word‐searches). Using multimodal conversation analysis, we show (a) that test takers avert their gaze from the screen (and hence from the coparticipant) in the course of trouble and revert gaze toward screen/coparticipant at trouble resolution; and (b) that these gaze behaviors differ in respective proportion and length across proficiency levels. The findings offer new insights into L2 IC assessment, gaze, and video‐mediated interaction research.