Exploring the Correlation of Physiological Stress Signals with Student Exam Performance: A Preliminary Study.

Abstract

Stress responses in real-world settings are less studied compared to controlled laboratory environments, limiting our understanding of their impact on cognitive performance. This study investigates the relationship between physiological stress signals and academic performance using an open-access dataset of 10 students assessed across three exam sessions (Midterm 1, Midterm 2, and Final Exam). Physiological measures, including electrodermal activity (EDA), heart rate (HR), and skin surface temperature (TEMP), along with exam grades, were analyzed using traditional hypothesis testing, bootstrap method, correlation analysis, and regression tree modeling. To address the small sample size, we validated traditional hypothesis test results with the bootstrap method, and both approaches were compatible. Hypothesis testing revealed no significant differences in physiological measures across exam sessions, supporting the null hypothesis. Grades differed significantly between the final exam and both midterms (p < 0.05). Stress fluctuations were also analyzed across three periods (beginning, middle, and end) for each exam, revealing temporal response variations. Correlation analysis showed a moderate negative relationship between EDA and HR (r = - 0.504, p < 0.01) and a weak positive relationship between EDA and TEMP (r = 0.417, p < 0.05), both intensifying during the final exam. Regression analysis explained 78% of the variance in grades (R² = 0.78), with regression tree modeling identifying lower skin temperature (< 28 °C) and higher EDA (≥ 0.19) as predictors of poorer performance. These findings underscore the interplay between physiological stress responses and academic outcomes, emphasizing the need for further research and interventions to support student success.