Two Essays on the Effect of Virtual Identiy on Consumer Decisions

This dissertation is aimed at examining the effect of virtual identity on consumer decisions. Consumers are increasingly expressing their identities through time on the virtual world of the Internet by communicating with others and exploring themselves. Accordingly, they start to develop a new identity associated with the virtual world, namely virtual identity, which is separate from their actual life identity. Hence, there is an urgent need for researchers to understand the implications of virtual identity on consumers' decisions and lives in general. Despite the attention and effort researchers put into investigating virtual identity, there are no existing works have systematically investigated the impact of virtual identity on consumer decisions. In addition, there are no established reliable and valid scales for assessing virtual identity. Therefore, the two essays of this dissertation intend to fill in these research gaps.

Essay 1 aimed at examining the effect of virtual identity on impulsive consumption. We propose that when individuals' virtual identity is salient, comparing to when their actual identity is salient, they tend to more likely engage in impulsive consumption. In addition, we propose that this is because when individuals' virtual identity is activated, comparing to when their actual identity is activated, they elicit more of their multiple possible selves. While their multiple possible selves are activated, they tend to induce conflict among the self-goals associated with those multiple possible selves. This self-conflict tends to deplete consumers' limited self-control resources. As a result, people will be more likely to engage in impulsive consumption. Through four studies, we tested this proposed relationship between virtual identity and impulsive consumption, the proposed mechanism, self-control resources, as well as a theory driven moderator, self-construal.

Essay 2 aimed at developing a reliable and valid measure for virtual identity. Integrating the literature of social identity theory and identity formation theory, we proposed four dimensions for the construct of virtual identity. First, we generated an item pool around the four theoretical dimensions for measuring virtual identity. Then the items were subjected to an expert panel assessment and four studies for refining and validating the scale. Specifically, the factor structure of the scales was determined by exploratory factor analysis and confirmatory factor analysis. In addition, the reliability was assessed by internal consistency and test-retest reliability. Finally, the validity is examined through discriminant validity, convergent validity, and predictive validity. In particular, virtual identity scale was compared with similar identity and self constructs (e.g., actual self, ideal self, and ought self) for analyzing discriminant validity. Convergent validity is analyzed via the correlation between different sub-constructs. Furthermore, for predictive validity, two studies were proposed to put virtual identity scale into test to see whether different level of virtual identity would lead to different level of impulsive consumption and risk taking.