Press the 'A' key every time you feel your heart beat for 10 seconds. We'll estimate your heart rate and predict how you'd perform on an exam right now.
Every student has experienced exam stress at some point. But what if we could quantify it? This project investigates the physiological responses of students during exams, exploring how changes in heart rate (HR), electrodermal activity (EDA), and blood volume pulse (BVP) relate to stress levels and performance.
To understand the connection between stress and academic performance, we utilize a Wearable Exam Stress Dataset, which records real-time physiological data from students across three major exams: Midterm 1, Midterm 2, and the Final Exam. The dataset includes the following key indicators:
Physiological data provides a real-time window into cognitive load and emotional responses during an exam. Here’s what the signals indicate:
The core of this study lies in answering a critical question: Does heightened stress correlate with lower exam scores? By analyzing patterns in HR, EDA, and BVP, we aim to uncover insights that may help students and educators understand the role of physiological stress in academic settings.
Explore the interactive visualizations below to see the data in action!
About this visualization: Our analysis of heart rate data shows that students with more stable heart rates throughout the exam tend to score higher, suggesting consistency in physiological response benefits performance. And our visualization reveals a surprising twist: the students with the highest peak heart rates early in the exam often experience a sharper decline mid-exam, rather than a steady stress adaptation. But this challenges the assumption that early nervousness dictates final outcomes, as some students recover and still perform well. Therefore, understanding individual heart rate patterns and managing fluctuations—rather than just reducing initial anxiety—may be key to optimizing test performance.
Select a test to begin visualization.
About this visualization: By visualizing students' heart rates during exams, we see that sustained high stress often correlates with lower exam performance, suggesting stress may negatively impact cognitive function. But our data also highlights individual exceptions—some students perform well despite elevated heart rates—indicating stress isn't the sole predictor of outcomes. Therefore, while a higher heart rate usually relates to poorer performance, our visualization emphasizes that stress impacts each student differently, reinforcing the complexity of exam performance.
About this visualization: By visualizing students' electrodermal activity (EDA), we clearly observe stress through measurable sweat gland responses. But not all students display identical patterns or timing of these stress signals, suggesting individual differences. Therefore, although higher EDA generally correlates with heightened arousal, our visualizations highlight that stress is multifaceted and manifests uniquely across students.
Select a test to begin visualization.
About this visualization: By visualizing average heart rate (HR) and electrodermal activity (EDA) over time, we observe clear patterns of stress throughout the exams. And by incorporating a 2D polynomial regression model, we can effectively predict exam scores based on these stress signals. But the regression also reveals that higher stress indicators don't uniformly predict poor performance—some students manage higher HR and EDA levels without significant performance drops. Therefore, our visualization and regression emphasize that the connection between physiological stress and exam outcomes is nuanced, highlighting that stress alone does not define academic performance.
Even though people commonly believe elevated stress signals—like a rapid heart rate—negatively impact test scores, our data shows that stress does not automatically mean poor performance. In fact, our visualizations reveal that stress signals can correlate strongly with improved exam scores, highlighting that the relationship between stress and performance is nuanced and complex. Other factors, such as preparedness and effective time management, also significantly influence exam success. Our visualizations clearly emphasize this complexity, demonstrating that the relationship between stress and performance is far from straightforward.
By combining realism (like pulsing hearts) with layered data (scatter plots, line charts, regression sliders), we keep the experience both emotional and informative. Viewers can discover a new perspective on stress—one backed by data and, hopefully, a bit of empathy.
In the end, stress is complex. Our biggest lesson is that being aware of one’s physiological responses can spark better coping strategies. Yes, stress may actually correlate with performance increases, but many variables play a role, and we hope this project inspires students to take ownership of their stress.