Assessing dogs’ performance in a social and non-social reversal learning task

Abstract

Reversal Learning could be an essential tool for dogs to accomplish a favorable adaptation to the human environment. Some dimensions of the social context, such as the presence of humans as choice stimuli, might influence dogs’ achievement in reversal learning tasks. Our goal was to assess the influence of the human presence on dogs’ ability to solve these tasks. For that purpose, we compared the performance of the same subjects in a social and non-social condition. Dogs had to choose between two passive humans (social reversal task) and between two apparatuses (non-social reversal task) as the discriminative stimuli. Our results showed no significant differences in the mean number of trials before giving the first correct response and mean number of correct responses comparing the social and non-social reversal conditions. This could indicate that reversal learning is independent of the social nature of the acquired stimulus, and that the human presence might not facilitate dogs’ performance. However, in the last block of trials, dogs made significantly more correct responses in the social task than in the non-social task. This result must be considered with caution. Further research is required to compare social and non-social tasks applied to the same subjects and including distinct dimensions of the social context. In addition, future work should address other factors that potentially shape dogs’ ability to learn reversals.