Study 1

Twenty-eight participants (50% female participants, 14/28) were recruited for the experiment (mean age 23.89 years, SD 3.08 years). One participant was excluded from the analysis because of having a high risk of mental illness, as assessed by the Symptom Checklist-90-revised (SCL-90-R; mean score 34.07, SD 40.22) [22]. The experimental protocol was approved by the institutional review board of the research site university.

The virtual environment of the experiment was implemented using Unity3D 2017.3.1f (Unity Technologies). The software was executed on a desktop PC with a Nvidia GeForce GTX 1080 graphics card (NVIDIA). The participants wore a head-mounted display (HMD; Oculus Rift CV1, Oculus VR) with a resolution of 1080×1200 pixels per eye and a refresh rate of 90 Hz.

A motion-capture system (Motive 2.0.2; NaturalPoint) was used with 14 Flex13 cameras (NatrualPoint) and 37 markers on a motion-capture suit. Each Flex13 camera had a resolution of 1280×1024 pixels, a frame rate of 120 frames per second (fps), and a latency of 8.33 ms. The positions of the markers within the tracking area were mapped onto a predefined avatar skeleton using Motive 2.0.2. After generating the avatar skeleton, the values for the skeleton bones with six degrees of freedom were streamed using socket communication at 120 fps with a streaming latency of less than 3 ms.

For this study, the following three avatar conditions were implemented: HA, PLA, and OBPLA.

The HA condition involved moving a human-shaped avatar using a full-body motion capture system. Three Unity3D modules were developed to implement the HA condition. The first module was an avatar animator module, which streamed 21 bones of the skeleton in real time from Motive 2.02 to the avatar. The second module was a size adjuster module, which matched the size of the avatar to the actual participant’s body size using the height, shoulder, waist, and pelvis sizes of the participant measured before the experiment. The gender of each avatar was also matched to that of the participant. The third module was a head rotator module, which received the head rotation information of the HMD and rotated the camera in the virtual space, providing visual information to the participant.

The PLA condition involved moving a PLA using a full-body motion capture system. The PLA was composed of 15 spheres, each with a diameter of 5 cm, according to the study of Troje [21]. The 15 spheres were all white without shadows. In the PLA condition, the avatar animator module sent skeleton data to the PLA, and different from the HA condition, the size adjuster module reflected the height, shoulder, and pelvis sizes. The head rotator module was operated the same as under the HA condition. Therefore, the camera in the virtual space rotated and moved to match the head position of the PLA.

The OBPLA condition utilized the out-of-body experience in the PLA condition, so that the PLA could be seen from the outside. The OBPLA condition was created by moving the camera in the PLA condition backward by 120 cm and upward by 30 cm relative to the global coordinate axis. The participants were able to observe their PLAs from a third person perspective.

A virtual mirror was placed in the virtual room so that the participants could easily observe their avatar. Participants saw the avatar by both looking directly toward the avatar and by looking at its reflection in the mirror. Several objects, such as computers, a copier, and chairs, were placed in the room for participants to be able to perceive their own body size in relation to these objects.

As part of a prequestionnaire, personal information, such as gender and age, was obtained, and the SCL-90-R was used. An embodiment questionnaire (EQ) was used to measure the level of the SoE. A part of the EQ developed by Piryankova et al was selected and modified to fit our experiment [23]. The questions in the EQ are explained in detail in Table 1. The EQ used a seven-point Likert scale to measure agency, ownership, and self-location separately. Participants filled out the EQ after experiencing each condition.

List of the items used in the embodiment questionnaire.

^aThe scale ranges from 1 (fully disagree) to 7 (fully agree).

Prior to the experiment, all participants completed consent forms and were instructed on the experimental procedures. All participants also completed a set of prequestionnaires. After the prequestionnaires were completed, participants watched a video clip containing human motion for 5 minutes to prevent participant immobility. Thereafter, participants wore a motion-capture suit and practiced the HA condition for 3 minutes to familiarize themselves with VR and the motion-capture environment. In the main session, participants were asked to move freely in the virtual room under the three conditions. Each condition lasted for 5 minutes and was counterbalanced. After completing each condition, participants removed the HMD and completed the EQ. The purpose of the study was explained to the participants once they had completed all tasks.