Effect of Response Delay Distribution on Sense of Agency and Performance

(Modeling using Free Energy and Validation Experiment)

Abstract

Sense of agency is the sense that I am the one who is causing or generating an action. Sense of agency is a factor of the pleasure of operating a machine. In recent years, machines have become increasingly autonomous. While machine autonomy increases task performance, it decreases the sense of agency. It is useful to construct a mathematical model of sense of agency and investigate the effect of the model variables on performance to balance sense of agency and performance. The mathematical model that we constructed predicts the interaction between prediction error and prediction uncertainty in both sense of agency and performance. In a simplified system of an automobile, prediction error was manipulated using the difference between the average of delay in learning and the average of delay in measurement, and prediction uncertainty was manipulated using the standard deviation of delay in learning. The interaction between prediction error and prediction uncertainty was examined in the measured values of sense of agency and performance. This interaction means that when an unexpected delay occurs, both sense of agency and performance are higher when one is not accustomed to the previous delay.