Feedback of action outcome retrospectively influences sense of agency in a continuous action task.

Hiroyuki Oishi,Katsumi Watanabe,Kanji Tanaka

doi:10.1371/journal.pone.0202690

Hiroyuki Oishi, Katsumi Watanabe + Show 1 more

Open Access

https://doi.org/10.1371/journal.pone.0202690

Copy DOI

Abstract

Here, we investigated whether explicit feedback on the result of the action (success or failure) modified sense of agency (SoA) in a continuous action task. Participants carried a white dot with a delay to a target square while avoiding obstacle squares. The color of the target changed unpredictably between white and blue. A trial was considered as successful or failed if the dot reached the target while it was white or blue. Thus, actions during the task resulted in almost identical experiences of successful and failed trials. After each trial, the participants reported to what extent they felt that they had been in control of the dot. The results showed that SoA was higher with shorter delay (i.e., easier control) and in the successful trials. These findings indicate that the sense of online control and the evaluation of continuous action based on feedback independently influence SoA. Particularly, the evaluation retrospectively modulated SoA.

Highlights

Sense of agency (SoA) is a subjective feeling that one is controlling one’s own actions and one’s action causes changes in external circumstances [1,2]
We investigated whether SoA in the continuous task was influenced by feedback of the final outcome
We found that: (1) SoA was attenuated by longer delay between key press and response

Summary

Introduction

Sense of agency (SoA) is a subjective feeling that one is controlling one’s own actions and one’s action causes changes in external circumstances [1,2]. Numerous studies have shown that SoA can be augmented or attenuated by internal motor signals and/or external cues [3,4] with tasks involving a simple action [5,6] and a continuous action [7,8]. As an example of explicit measurement of SoA, after participants finished a given task, they answered to what extent they caused the external event by either verbal report or use of computer mouse [5,9]. The comparator model (a computational model of motor control) suggests that SoA basically arises from internal cueing and is determined by a match between a sensory prediction of an action’s consequence and the actual consequence [11,12]. If the prediction and the actual result match, the sensory event is likely to be recognized as generated by the agent, resulting in a strong SoA. If the predictions do not match, the event is regarded as an external

Methods

Results

Conclusion