Personality in Capuchin Monkeys (Sapajus apella): Comparison of Assessment Methods and Constructs

Understanding personality in nonhuman primates is critical for advancing theories of individual differences, yet methodological variation remains a major challenge. We examined whether three commonly used approaches-questionnaire ratings, behavioral observations, and experimental tests-capture stable and overlapping personality constructs in 26 captive tufted capuchin monkeys (Sapajus apella). Using Bayesian multivariate mixed models, we retained individual variability to assess temporal stability, cross-context consistency, and inter-method convergence. Behavioral observations revealed conceptually coherent clusters related to stress, activity, and social engagement, but temporal stability was low. Novelty Tests produced associations suggestive of approach-avoidance and activity-focus dimensions, yet these clusters also lacked temporal stability. Questionnaire ratings showed partial convergence with observational data but not with novelty test data. Out of 39 predicted associations, only 4 were confirmed, highlighting weak cross-method consistency. Housing conditions significantly influenced personality expression, whereas sex and age effects were minimal. These findings suggest that personality expression may be more context-sensitive and multidimensional than is captured by any single assessment method, and that divergence across methods may reflect different aspects of personality rather than measurement error. Integrative, multi-method, and longitudinal designs that incorporate multiple measures may help improve the interpretation of primate personality assessments.