Instructor presence in educational videos: A systematic literature review of eye‐tracking research on learning outcomes and gaze behaviour

Despite growing eye‐tracking research on instructor presence, findings have not been comprehensively synthesized. This systematic review examines eye‐tracking research on instructor presence in educational videos, synthesizing evidence from 51 studies to understand how instructors and accompanying variables influence gaze behaviour and learning outcomes. To identify the contributions of eye‐tracking methodology, we examine the prevalent eye‐tracking metrics employed and their interpretations, as well as how studies investigated the association between gaze patterns and learning outcomes. Our key findings are as follows: (1) eye tracking is used almost exclusively to investigate visual attention; (2) changes in visual attention do not reliably predict changes in learning: only eight studies directly tested this link using inferential statistics, results were mixed and context‐dependent, and mediation analyses generally did not support visual attention as the mechanism by which instructional conditions improve learning, cautioning against treating eye‐tracking measures as a standalone proxy for learning effectiveness; (3) while instructors naturally capture attention, deictic cues effectively redirect focus toward learning content; (4) instructor presence produces mixed results, with most studies reporting either non‐significant or positive effects on learning, and negative effects being rare; (5) instructors' use of pointing gestures and attentional cues consistently enhance learning, whereas guided gaze and affective cues show more context‐dependent effects; (6) learner prior knowledge emerges as a strong moderator, with consistent evidence that low prior knowledge learners benefit more from deictic cues and positive affective cues, while other learner attributes and material characteristics primarily influence learning and attention through interactions rather than main effects. Context and implications Rationale for this study: The pedagogical value of including a visible instructor in educational videos is contested, with theoretical perspectives divided. Eye‐tracking research can clarify how learners attend to instructors versus content, but the rapidly growing evidence base in this area has not been systematically synthesized across metrics, conditions and outcomes. Why the new findings matter: This review shows that instructor presence consistently attracts attention but produces mixed effects on learning, with specific instructional behaviours moderating both. Critically, visual attention may not reliably predict learning outcomes; this finding cautions against treating eye‐tracking measures as standalone proxies for the effectiveness of video‐based instruction. Implications for practitioners and researchers: When instructors are included in educational videos, peripheral positioning, pointing gestures, explicit attentional cues and naturalistic positive affect best support learning, particularly for learners with low prior knowledge. Beyond these design recommendations, the review highlights a methodological gap, as very few studies statistically test whether gaze patterns mediate learning effects. Future work should therefore employ mediation and moderation analyses to clarify whether attention shifts are the mechanism through which instructional manipulations improve learning, or whether such links depend on specific conditions, rather than inferring this relationship from descriptive comparisons alone.