Developing Skills in Virtual Reality: the Influence of Alien Motion on Sense of Embodiment, Self-Efficacy and Training Outcomes

The use of Virtual Reality for training purposes has shown to be highly effective. Training new skills or behaviours using this technology can lead to improved outcomes, engagement and motivation. Within a virtual environment, people can be supported through alien motion or false feedback without them being aware of it. These methods can alter the behaviour of virtual objects the user interacts with to make task completion easier. It's important to prevent such modifications from affecting the users' Sense of Embodiment, necessary to give them the illusion of being the ones executing the actions in the virtual scene. Many virtual training systems only focus on users' performance, without considering their psychological state. Self-efficacy, or confidence in one's capabilities, is a key factor for effective training. Experiencing success and avoiding repeated failures is essential to develop it. The current dissertation investigates the effectiveness of using alien motion to support users during a training process. We examined its influence on Sense of Embodiment, self-efficacy and training outcomes in terms of how many times and how accurately users achieve their goal. We developed a Virtual Reality training application for a ball-throwing task in which users have to hit the centre of a target. The system can support participants by enhancing the trajectory of the ball, without them being aware of it, to make them more confident about their abilities. Various support management methods were tested, including no support, static support and adaptive support, to understand which approach more positively influences performance and self-efficacy without affecting the Sense of Embodiment. According to our findings, supported users showed higher improvements in accuracy and number of successful hits. Furthermore, adaptive support prevented decreases in self-efficacy and Sense of Embodiment. This project was realized during my internship at the National Institute of Informatics in Tokyo, Japan, under the supervision of Professor Soccini and Professor Inamura.