Design and validation of a 360° augmented-virtuality classroom with embedded eye tracking and pupillometry measures within a chroma-key arena
摘要
A growing body of work uses virtual reality systems as experimental platforms to study human perception, cognition, and behavior. While these systems offer precise temporal control and experimental replicability, their dependence on computer-generated environments has been criticized for weakening external validity. This concern is especially salient in psychoeducational research, where behavioral responses are shaped by instructional and social context. Recent advances in mixed reality address this challenge by combining physical and virtual elements. In this study, we evaluate an augmented-virtuality classroom system that embeds a pre-recorded 360° video within a chroma-key arena, coupled with synchronized binocular eye-tracking and pupillometry measures. During a ~ 15-min lecture, data completeness was high (median valid-frame fraction 0.92, n = 68), with 66 participants meeting all analytic quality checks for saccades and pupillometry measures. Oculomotor analyses yielded median event rates of 47.6 saccades/min and 88.4 fixations/min. Pupil size was centered around 2.27 mm. Classroom distractors elicited reliable phasic dilation, quantified as area under the curve (AUC, mm·s). The left eye increased from − 0.0099 ± 0.0172 to 0.1271 ± 0.0264 (paired t-test, p = 3.15 × 10⁻⁶), and right eye from − 0.0162 ± 0.0172 to 0.0945 ± 0.0228 (p = 3.00 × 10⁻⁴). Latency measures placed peak pupil dilation at a median of 4.19 s in the left eye and 4.03 s in the right eye post-onset. Participants’ gaze across time clustered near the lecturer and podium, aligning with selective-attention patterns in classroom settings. These findings demonstrate that reproducible and controllable augmented-virtuality classrooms with high-quality oculomotor measurement can be used in psychoeducational investigations.