Statistical theories and their proofs may be immutable, but the rubber meets the road when theory turns into practice. In this talk, I will argue that, as the practice of statistics by humans is fundamentally a psychological action, we should take a psychological approach to researching the act of "doing statistics". This will be done through two examples: one project inspired by the cognitive sciences to study how graduate students process p-values, and the other inspired by the learning sciences to study how graduate students think about statistical tests. Through these examples, I hope to show how the study of the psychology of statistics can help statisticians ameliorate long-standing problems in practice and pedagogy.