Reflecting and Shaping the Self through Avatars: The Relationship between Avatars, Identity, and Personal Needs

Individuals frequently engage with virtual environments through the use of characters that represent the self, known as avatars. This dissertation focuses on two primary research questions: (1) how do avatars reflect identity and, (2) how does engaging with an avatar shape the self, in terms of personal needs and self-perceptions? We examine the bidirectional relationship between avatars and their users across four studies. Study 1 examines whether customized avatars can accurately communicate the personalities of their creators to others. Expanding on the theme of reflecting identity, Study 2 explores whether avatar preferences are related to individuals personal psychological needs, specifically the needs for warmth and competence. The results of Studies 1 and 2 indicate that avatars can accurately reflect identity in terms of both personality and psychological needs. However, individuals can also be motivated to use avatars in a way that deviates from ones actual identity, such as avatars that reflect ones ideal self. Study 3 examined whether creating an avatar provides individuals with the opportunity to self-enhance in response to psychological threat. Specifically, we investigated whether there is a tendency to create more idealized avatars following psychological threat and whether this can help mitigate the negative effects of threat on mood and self-concept. The results did not support these ideas, however, with avatar creation seeming to exacerbate rather than improve the negative outcomes of experiencing a psychological threat. That said, it is possible that actively controlling an avatar is an important prerequisite for avatars to have a positive influence on self-perceptions. In Study 4, participants were asked to create either an avatar that reflected their actual self or their ideal self; they were subsequently assigned to either watch or control this avatar. Controlling an avatar, regardless of type, was related to improvements in self-concept (e.g., self-liking), but did not any reduction in discrepancy between the actual self and ideal self. We discuss the results of these studies with a focus on how they might inform future work and their possible application in the real-world, including interactive social interventions.