Predicting sense of agency using smartphone and wearable device logs

Abstract

Sense of agency (SoA) is the subjective experience that one’s volitional action caused an event. It has been widely investigated that SoA plays a relevant role in shaping and moderating human cognition, emotion, and behavior. Thus far, the common ways to quantify SoA are psychological scales and cognitive tasks in laboratory settings, while no efficient real-time measure has been proposed, which limits the scope for investigating and applying real-time SoA in a real-world setting. To address this issue, we investigated the feasibility of estimating SoA from daily behavior logs (e.g., heart rate, physical activities, environment illumination, and smartphone-usage time). Using a smartphone app and wearable devices, we collected behavioral logs and the grand truth of SoA (n=43, over 600 data points). Using this dataset, we built and compared multiple models to estimate the SoA level. The results showed that models with daily behavior logs as independent variables significantly outperformed the baseline model, confirming the predictive utility of daily behavior logs for SoA..