# Predicting Performance Under Stressful Conditions Using Galvanic Skin Response

**Authors:** Carter Mundell, Juan Pablo Vielma, Tauhid Zaman  
**arXiv:** [1606.01836](https://arxiv.org/abs/1606.01836)  
**Date:** 2016-06-06  
**Summary page:** https://www.summarizepaper.com/en/arxiv-id/1606.01836v1/

## Key points

- Rapid growth of wearable biosensors enables measurement of worker performance using biological signals
- Challenge lies in predicting performance under stressful and high-risk conditions
- Researchers focus on galvanic skin response (GSR) as a potential predictor of worker performance
- Experiment conducted with subjects answering arithmetic questions under low and high stress conditions, monitoring GSR using wearable biosensor
- GSR data collected during low stress conditions successfully predicted performance under high stress conditions (AUC value of 0.76)
- Predictions without biometric signals had AUC barely exceeding 0.50
- Wearable biosensors can provide valuable insights into predicting performance in high-stress scenarios based on data from low-stress situations
- Biometric signals can optimize worker performance and enhance safety measures in challenging work environments
- Organizations can develop proactive strategies by understanding how biometric signals correlate with performance outcomes
- Leveraging wearable biosensors and analyzing specific biological signals like GSR offers predictive capabilities for assessing worker performance under stressful conditions

## Summary

The rapid growth of wearable biosensors has opened up new possibilities for measuring worker performance using biological signals. However, the challenge lies in not just measuring performance, but actually predicting it under stressful and high-risk conditions. In this study, the researchers focus on one specific biological signal known as galvanic skin response (GSR) and investigate its potential for predicting worker performance. To test their hypothesis, the researchers conducted an experiment where subjects were asked to answer arithmetic questions under both low and high stress conditions. Throughout the experiment, the subjects' GSR was continuously monitored using a wearable biosensor. The results of the study revealed that by analyzing only the GSR data collected during low stress conditions, it was possible to predict which subjects would perform well under high stress conditions. The prediction accuracy achieved an area under the curve (AUC) value of 0.76. In contrast, when attempting to make similar predictions without utilizing any biometric signals, the AUC barely exceeded 0.50. These findings suggest that wearable biosensors can provide valuable insights into predicting performance in high-stress scenarios based on data obtained from low-stress situations. This research highlights the potential applications of biometric signals in optimizing worker performance and enhancing safety measures in challenging work environments. Organizations can leverage these insights to develop proactive strategies that support workers and mitigate risks associated with high-stress tasks by understanding how biometric signals correlate with performance outcomes. Overall, this study demonstrates how leveraging wearable biosensors and analyzing specific biological signals like GSR can offer predictive capabilities for assessing worker performance under stressful conditions. By understanding how these signals correlate with performance outcomes, organizations can potentially implement proactive strategies to support workers and mitigate risks associated with high-stress tasks.