Abstract # 960 Event # 118:

Scheduled for Friday, August 19, 2005 09:45 AM-10:05 AM: Session 9 (Crystal Ballroom) Oral Presentation

How should personality be measured in primates?

S. D. Gosling1, S. Vazire1, A. S. Dickey2 and S. J. Schapiro3
1Department of Psychology, 1 University Station A8000, University of Texas, Austin, TX 78712, USA, 2University of Iowa, 3University of Texas, M. D. Anderson Cancer Center
     To lay the conceptual groundwork for the symposium on primate personality, we explore the core measurement issues underlying all attempts to assess personality in nonhuman animals. Assessments of animal personality have faced three concerns: 1) that personality cannot be measured reliably in animals, 2) that assessments of animal personality are overly subjective, and 3) that the methods required to obtain valid personality assessments are impractical. Using new data from a captive colony of chimpanzees (Pan troglodytes) along with previous findings in the human and animal literatures, we address each of these concerns and evaluate the viability of primate-personality assessments. We assessed 52 group-housed chimpanzees (at the Department of Veterinary Sciences of The University of Texas M. D. Anderson Cancer Center) using both coding and rating methods. For the codings, subjects were observed for eight to twelve 15-minute focal-animal samples as they freely interacted within their social groups. For the ratings, four observers who knew the animals well rated the individuals on a list of 34 behaviorally defined traits. The magnitude of the coding-rating convergent correlations varied dramatically (rs = -0.52 to 0.59). Reliabilities were generally stronger for ratings (mean ICC = 0.61) than for codings (mean ICC = 0.42). Additionally, conceptual analyses showed that ratings are not as “subjective” and codings not as “objective” as is often assumed. Together, findings favored the use of rating methods as measures of personality.