Two often disparate bodies of literature have established the existence of two motivations: positivity and consistency. Positivity motivation often manifests as illusory self-enhancement, as demonstrated by the better-than-average effect and self-serving attributions. Consistency motivation has been established in the empirical traditions inspired by balance theory, cognitive dissonance theory, and self-verification theory. Given the robust literatures supporting these two motivations, this dissertation attempts to identify moderators, such as the privacy of the situation and trait preference for consistency, that determine under what conditions and for whom each motive will take precedence. In addition, I attempt to integrate various consistency theories by proposing that the same basic consistency motive is responsible for cognitive dissonance, balance, and self-verification effects. In Study 1, participants encountered a norm supporting or undermining norms about consistency, believed that they were participating publicly or in private, and then chose between a partner who viewed them positively and a partner who viewed them consistently with their own self-concepts. Privacy, norms, and individual differences did not influence participants' choices, disconfirming all hypotheses. In Study 2, the same moderators (norms, privacy, and individual differences) were applied to a self-affirmation paradigm. After encountering either a pro-consistency or anti-consistency norm, female participants self-affirmed publicly, privately, or not at all, and then completed a math test under stereotype threat conditions. Norms, privacy, and personality did not influence performance on the math test, failing to confirm hypotheses and replicate previous research. I consider a number of methodological explanations for these null results working under the assumption that these null results are a Type II error. However, none of the explanations considered fully explain the findings, suggesting that moderators of dissonance effects may not generalize to other consistency paradigms.