Proceedings of the International scientific and practical conference ―Synergy of Modern Science and Education‖ (February 2-4, 2026) / Publisher website: www.naukainfo.com. – New York, USA, 2026. - 324 p.

72 studies on this topic established fundamental regularities between pupil dilation and cognitive load. Recent reviews confirm the suitability of this indicator for measuring cognitive effort in learning contexts [6][7][8];  Blinking (blink rate) is often considered a proxy for load or fatigue. Studies on blink rate and workload, including research in driving contexts, suggest a relationship between blink rate and workload, although findings in the literature are mixed [9]. Practically, this means blink metrics are better used as auxiliary, not as the sole indicator. It is also important that research on the applicability of webcam-based eye tracking in learning scenarios is emerging. Webcam-based gaze tracking has been demonstrated to reach accuracy that can be acceptable for certain educational tasks [10]. Key limitations of the webcam-based approach:  Accuracy is significantly lower than hardware eye-trackers (1–2° error vs 0.5°);  It works better for tracking general regions of attention than precise fixations.  It requires calibration for each user;  It is sensitive to lighting, posture, and camera quality. Nevertheless, for the task of detecting overload through changes in gaze patterns (rather than exact coordinate tracking), this approach is promising. Behavioral interaction traces can serve as ―state sensors‖ without requiring a camera, while ocular signals from eye tracking can further enrich state inference when available [11]. Cues of overload include:  Increased time per step/screen;  Frequency of errors in exercises/tests;  Backtracking (frequent returns);  Increased number of attempts;  ―Chaotic clicks,‖ mis-clicks, frequent opening/closing of panels;