Multimodal Analysis of Performance Resilience Under Increasing Cognitive Load Using Eye-Tracking and Behavioral Measures

This paper presents a multimodal approach to analyzing performance resilience under increasing cognitive load using a Digit Symbol Substitution Test (DSST) combined with eye-tracking measures. A total of 61 participants completed three progressively challenging DSST conditions under time pressure, during which eye movements, pupil dynamics, and behavioral performance metrics were recorded. Data from 57 participants were included in the analyses. Cognitive load was increased by systematically varying task complexity and duration across conditions. The analysis focused on relationships between eye-tracking features and behavioral performance, as well as on individual variability in response to increasing task demands. Linear regression was used to characterize consistent trends across participants, including changes in behavioral performance and eye-movement patterns with higher task demands. Importantly, a subgroup of participants exhibited stable or improved performance despite increasing cognitive load, indicating performance resilience. While their behavioral outcomes remained stable, subjective and eye-tracking measures indicated that task demands increased across the sample, suggesting a dissociation between maintained performance and experienced demand. These findings demonstrate that combining eye-tracking and behavioral measures enables the identification of individual differences in performance resilience under cognitive load. The proposed approach provides a basis for non-invasive assessment of cognitive functioning and may support applications in cognitive monitoring and mental fatigue research.