Reliability of a method to measure novelty-seeking in nonhuman primates

The objective of this study was to create a reliable method to measure novelty-seeking (NS) temperament in stumptailed macaques. We correlated two behavioral indexes (risk-taking and curiosity) with the NS index. Cochran's concordance index yielded ethogram reliability (risk: alpha = 0.88; curiosity: alpha = 0.79). Risk and curiosity indexes were obtained with a synthetic index obtainment approach. Intraclass correlation of the instrument yielded a value of 0.88. Temperament index was established from the average score given by each judge. The instrument presented 92% specificity and 70% sensitivity, with a 56-point score taken as the cuttoff point to classify individuals. The results of a Spearman's analysis of the three indexes were significant (temperament-risk, n = 32, P<0.01; temperament-curiosity, n = 29, P<0.05; risk curiosity, n = 29, P<0.05). We conclude that the method we developed is reliable.