Does the sex of the human demonstrator influence the performance of dogs during an observational learning task? Preliminary results

DZIK, M.V.; GUTIERREZ, S.; BERDUGO-LATTKE, M.; BENTOSELA M.

Congreso; VIII CANINE SCIENCE FORUM; 2023

Introduction. Dogs can learn to perform actions through the observation of human demonstrators. Moreover, they behave differently when interacting with men or women. While they present more stress-appeasement behaviors towards men, they may prefer women due to more positive verbalizations and gentler pets.Question/Hypothesis. The objective was to evaluate the effect of demonstrator’s sex on observational learning. It would be expected for dogs with female owners to perform better.Method. The sample comprised 43 adult dogs, of both sexes and various breeds. Dogs had to learn to open the lid of a container using their paw or snout to obtain food. Dogs from the experimental condition (N = 29) observed their owner opening one of two containers, while dogs from the control condition (N = 14) received no demonstration and observed the baited container already opened. The owner was male for half of the dogs on each condition, and female for the other half. There were two phases of four demonstration trials and one test. The evaluation was carried out virtually through Zoom.Results. Generalized Linear Mixed Models were used. Compared to control, in the experimental condition dogs chose (F (1,19) = 318.18, p < 0.001) and opened (F (1,32) = 5.83, p = 0.022) the target container more frequently. Likewise, they exhibited greater contact with the target container in the first test (condition *phase: F (1,41) = 7.10, p = 0.011), and less contact with the non-target container in both tests (1,12) = 9.15, p = 0.011). In addition, dogs in the experimental condition looked longer at the target container’s area during the demonstration phase (F (1,23) = 79.36, p < 0.001). Finally, dogs with female owners chose the target container more frequently in the experimental condition than the control (condition*owner's sex: F (1,14) = 224.70, p < 0.001). In contrast, male-owned dogs chose the correct container more and looked longer at the target container’s area during the demonstration phase than female-owned dogs (F (1,14) = 256.56, p < 0.001), regardless of the condition.Discussion. Dogs were able to learn to solve the task by observing a human. Only those with a female owner benefited from the demonstration, showing a moderate effect of the demonstrator's sex on said learning. These findings are relevant in applied settings and for the selection of research samples, as female owners are often more prevalent.