Reaffirming the validity of unobtrusive personality measures: A comment on Harms et al. (2025) and guidance for the field

Abstract

As unobtrusive measures of CEO personality have become more common in organizational research, questions about their validity have grown. Harms et al., Journal of Occupational and Organizational Psychology, (2025) 98, e70052 recently formalized some of these concerns in a meta-analysis, citing higher average intercorrelations and weaker associations with firm performance relative to survey-based measures as evidence of their lack of validity. We argue that these criticisms are overstated, reflecting both a misunderstanding of what unobtrusive measures capture and limitations in the authors' assumptions and meta-analytic approach. Whereas self-report surveys capture executives' self-concepts, existing unobtrusive measures typically assess externally observed or expressed personality. We suggest that relatively higher average intercorrelations are theoretically expected when assessments reflect stylized expressions of CEOs' observed tendencies, and we provide evidence that unobtrusive measures can meaningfully distinguish between traits. We further argue that, given what they represent and their application in large firms, unobtrusive measures likely relate differently to outcomes shaped by external audiences as compared to aggregate financial outcomes. Consistent with this logic, we show substantive differences in the relationships between unobtrusive measures of CEO personality and market performance compared to accounting performance. Our study reaffirms the validity of unobtrusive personality measures and offers guidance for their appropriate evaluation, interpretation and continued refinement.