Toward Mental Effort Measurement Using Electrodermal Activity Features.

Area of Science:

• Physiological computing
• Human-computer interaction
• Wearable technology

Background:

• Monitoring mental effort is crucial for optimizing productivity in work and study.
• Electrodermal activity (EDA) is an emerging physiological signal for assessing mental workload.
• Wearable sensors offer a non-invasive method for continuous physiological data collection.

Purpose of the Study:

• To evaluate the efficacy of electrodermal activity (EDA) features in predicting self-reported mental effort.
• To explore the use of interpretable machine learning for mental effort monitoring.
• To assess the potential of wearable sensors for real-time productivity enhancement.

Main Methods:

• Analysis of over 92 hours of EDA data from a single participant across 91 diverse activities using the Empatica E4 wearable sensor.
• Feature extraction focusing on EDA signal intensity, dispersion, and peak intensity.
• Implementation of a logistic regression model to predict high mental effort states.

Main Results:

• EDA features related to signal intensity and peak intensity were most predictive of high mental effort.
• Increased EDA signal and peak intensity correlated with higher self-reported mental effort.
• The model achieved moderate predictive efficacy (AUC = 0.63, F1 = 0.63) when cross-validated, outperforming model bias alone.

Conclusions:

• Electrodermal activity (EDA) shows promise as a physiological indicator for sensor-based self-monitoring of mental effort.
• While EDA is a valuable indicator, integrating other physiological data (heart rate, respiration) may improve prediction accuracy.
• This research contributes to the development of technologies for enhanced productivity through physiological monitoring.