Measuring Skin Temperature as an Indicator for Cognitive Flow Using a Wrist-worn Wearable

As traditional fitness sensors have revolutionized engagement in exercise through continuous feedback on physical effort and exertion, continuous feedback provided by wearable sensors can help us optimize our mental engagement and experience. To that end, our research focuses on developing a framework for measuring cognitive flow using physiological data which can be readily collected through wrist-worn wearable sensors. Previous research suggests that movement and skin temperature are strong predictors of flow state. While movement is a precursor to flow in many cases, an increase in skin temperature is an outcome of being in flow and may be a robust predictor for flow across many types of activities. In this study, we compare skin temperature trajectories between activities which are conducive to flow and those which are not conducive to flow and evaluate the efficacy of the trajectory curve parameters in predicting flow. In this case study, we collected data from a single participant across 79 unique activities. Based on visual inspection of the trajectories, the polynomial family of models was employed, and we found that a third-order polynomial provided the best balance between fit and parsimony. The trajectory parameters and r-square fit index for each activity resulted in predictions of moderate accuracy (AUC > 0.7). Statistical models suggest that being in flow is indicated by a positive skin temperature trajectory. In future research, we hope to apply this model across multiple individuals within an expanded scope of activities. Such a tool can be used to assist individuals with different abilities and needs in achieving mental flow for learning tasks.