The Strategic Use of Instructor Visibility in Educational Videos: An Eye-Tracking Data-Based Study

This study developed an exploratory heuristic decision-support framework for dynamically regulating instructor visibility in instructional videos based on learners’ eye-tracking data. Eye-movement data were collected from 40 university students using a Tobii Pro Fusion while viewing 16 explanatory and procedural instructional videos. Instructor and Learning Content Areas of Interest were analyzed using normalized eye-tracking indicators. Descriptive analyses were complemented by Wilcoxon signed-rank tests, binary logistic regression, and receiver operating characteristic (ROC) analyses. Learners allocated significantly greater visual attention to learning content than to the instructor across both video types. Logistic regression identified instructor attention percentage, visit-related metrics, and blink frequency as significant predictors of visibility decisions. ROC analysis showed that instructor attention percentage provided the highest discriminative performance, with an optimal threshold of approximately 30%. Findings suggest that instructor visibility may be particularly appropriate during introductory, summary, and conceptually demanding segments, providing empirically grounded, attention-aware guidance for adaptive instructional video interfaces.

<br class="Apple-interchange-newline">