Procedure

The experiments were conducted in a dark and soundproof room. The experimenter and participant faced each other across a desk (60 cm apart) on which an LCD monitor was placed (see Fig. 5a). On the participant’s side, the platform where the mouse would be used (59 cm in height) was set under the desk (see Fig. 5a); therefore, the participant could not see either the mouse or his or her own hand. A guide rail and right and left stoppers were set to help the participants move the cursor along the long axis of the monitor screen (see Fig. 6). On the experimenter’s side, a stage for moving the mouse was set along the length of the monitor screen (width: 9.0 cm, height: 5.7 cm). On the experimenter’s side of the monitor screen, a guide image hidden from the participants’ sight by a low screen, shown in Fig. 5b, was displayed for the experimenter. The experimenter performed actions similar to those of the participant by following this guide image with his own mouse. The participants were clearly informed that the movements of the cursor on the monitor screen were caused by their own mouse movements or by prerecorded data on participants’ movements during the practice period. Thus, the participants knew that the movements of the experimenter’s mouse never affected the cursor movements. On the participant’s side, a head rest and a chin rest were positioned to hold the participant’s head still. (The head rest is shown in Fig. 5a, but the chin rest is not shown in the figure to avoid complicating the drawing). During the experiment, each participant placed his or her chin and forehead on the chin rest and head rest, respectively. Using the chin rest and head rest, the distance between the center of the screen and the center of the participant’s eyes was fixed at about 42 cm, and the angle of the participant’s view to the horizontal axis was fixed at about 50 degrees. A screen board was set under the desk to hide the movements of the participant’s arm from his or her own line of vision. Therefore, the participants were not able to see their own hand or arm movements during the task. They were also asked to put on headphones with white noise playing to mask the sound generated by the movement of the mouse on the platform.

First, the participants underwent one set of trials with the seven time lags for each condition as practice. During this practice period, the participants learned to move the mouse from left to right and from right to left following the beat of a metronome. The order of three conditions was fixed as follows in the practice period:

Self condition → Other condition → Both condition

For the Self condition in the practice period, the movement data from the participant’s mouse were recorded on a PC and later used to move the cursor automatically in the Other condition. We determined the values of the time lag and the pitch of continuous mouse movements to avoid movements with half-cycle differences between the participants’ mouse and the cursor because such reversals could also affect the sense of agency⁴⁵. Then, the main experimental rounds were conducted after the practice period. The order of the three conditions was counterbalanced among the participants. Intersession breaks of about 10 minutes were given between each condition. The participants performed eight sets of trials consecutively for each condition. One set of trials consisted of the seven time-lag values. The order of the time lags in the trials was randomized. In total, each participant performed 56 trials for each condition (168 trials in total).

The participants were asked to make a forced choice from four possible answers to evaluate their agency attribution. The prompt and four answer options were displayed on the monitor screen after each trial. The answers were recorded by pressing a key. The four answers were displayed at the corner of the rectangle, and the participants pressed the direction key corresponding to the positions of the answers on the monitor screen. The positions of the answers and keys were changed after every predetermined number of trials. Each trial for each time lag was executed in the order shown in Fig. 7.

Order of the procedures for each trial for each time lag. (1) The word “Start” was displayed on the monitor screen to instruct the participants to start moving their mouse. (2) The participants moved their mouse to control the cursor in the Both and Self conditions, and the cursor moved automatically in the Other condition. The experimenter moved his mouse using similar actions in the Both and Other conditions. (3) The participants stopped moving their mouse when a beep sounded after the mouse had been moved five times to the right and left in the Both and Self conditions, and the cursor stopped automatically after moving five times to the right and left in the Other condition. (4) The question and answers were displayed on the screen. (5) The participants selected the answer that best matched their feeling from the four options provided by pressing the key.