Proteus, the origin story: how avatar appearance in virtual environments can shape user behavior

March 2023
Learning and Innovating
Communications par affiche dans un congrès international ou national
Auteurs : Anna Martin (LINEACT), Beatrice Biancardi (LINEACT), Stéphanie Buisine (LINEACT)
Conférence : Interdisciplinary College 2023 Dynamics of Experience - Mind, Bodies and Things, 11 March 2023

Since their emergence, virtual environments have represented a new opportunity to study human behavior in situations impossible to replicate in the real world. One such situation is the embodiment of a virtual body entirely different from our own. Research in this area has revealed that the physical characteristics of avatars can elicit specific behaviors from the users embodying them, a phenomenon termed the Proteus effect. For example, a recent study has shown that embodying avatars resembling inventors in a brainstorming task increased creativity compared to non-inventor avatars. Even though many studies have demonstrated the Proteus effect in various contexts, very few have focused on its theoretical explanation. Therefore, the first goal of my PhD project is to investigate the cognitive mechanisms that underlie the Proteus effect. For that, my first experiment aims to investigate the effect of deindividuation on the Proteus effect. Deindividuation describes a psychological state in which a person’s behavior is driven by situational norms instead of personal norms. It can explain unusual behaviors observed in groups or under anonymity. It has been hypothesized that embodying an avatar induces a state of deindividuation which could facilitate the emergence of the Proteus effect. To test this hypothesis in our experiment, we plan on manipulating the state of deindividuation between participants (i.e. deindividuated or self-aware) and measure whether it moderates the occurrence of the Proteus effect. Since the second goal of my PhD is to find innovative applications of the Proteus effect in educational and professional contexts, we choose to study the effect of avatar height (i.e. short or tall) on participants’ leadership style (i.e. transformational leadership score). In the deindividuated condition, we expect tall avatars to be associated with higher transformational leadership scores compared to short avatars. However, if deindividuation moderates the effect, we don’t expect to find a difference between height conditions for the self-aware group. All in all, this project aims to deepen our knowledge about the complex interactions between avatars and their users as virtual environments and immersive technologies become increasingly more common.