No Compelling Evidence that Self-Reported Personality Traits Explain Basal Testosterone and Cortisol’s Associations with Status-Relevant Behavior

Abstract

Objective: A goal of behavioral neuroendocrinology is to understand how basal hormone levels relate to behavior. Studies of human participants sometimes measure self-reported personality traits, in addition to or instead of direct behavioral observation. Although personality traits often predict their respective behaviors, whether personality explains hormone-behavior relationships remains unclear. Methods: We obtained data from eight previous studies (total N = 985) that examined baseline testosterone and cortisol as predictors of status-relevant behavior (competitiveness, dominance, risk-taking, aggression, affiliation, and social status). We tested whether the previously reported hormone-behavior relationships are mediated by self-reported personality traits (e.g., trait dominance, prestige, extraversion). As a secondary research question, we also tested whether trait dominance moderated the testosterone-behavior relationships. Results: As expected, self-reported personality traits often predicted status-relevant behaviors, but there was little evidence that traits also correlated with basal testosterone or the testosterone × cortisol interaction. Across all eight studies, personality traits did not significantly mediate hormone-behavior relationships. Indeed, the effect sizes of the hormone-behavior relationships were robust to the inclusion of personality traits as covariates. Further, we did not find strong or consistent evidence that trait dominance moderates the testosterone-behavior association. Conclusion: Results suggest that basal testosterone and cortisol predict status-related behavior independent of self-reported personality. We discuss how these results may have broader implications for the physiological mechanisms by which testosterone and cortisol influence behavior, a process that could be unconscious and automatic. We also discuss alternative explanations, limitations, and future directions.