Observational learning and social facilitation in C57L and C3H strains of the mouse, Mus musculus.

The experiments described in this thesis investigated observational learning processes in C57L and C3H pure inbred mouse strains. Most of these experiments explored the modelling of actors by reactors in a 2-choice food-source discrimination situation where actors' food-source preferences were induced by electrifying one of the two food basins. The results of Experiment 1A were that with male-female mouse pairs of both strains, reactors modelled the food-source preferences of actors when in the presence of, and after exposure to, the discrimination learning of actors. The hypothesis that modelling processes were a major determinant of reactors' preferences gained support from the results both in the control groups and in Experiment 2 (concerning the role of olfactory deposits) and in Experiments 3 and 9 (influences from social facilitation processes). Also, the results of Experiment 4 suggested that ultrasonic signalling played a minor part in the social behaviour sampled in Experiment 1A. However, in the replication of Experiment 1A with male-male and female-female pairs as subjects (i.e. Experiment 1B), modelling in reactors was far less evident. In subsequent experiments the behaviour observed in Experiment 1 was further analysed by studying modelling in two basic situations, namely, (a) where reactors were tested in the presence of trained actors, and (b) where, after exposure to actors' food-source preferences, reactors were tested in the absence of actors. The findings of Experiment 5 suggested that male-female pair reactors from both strains would model when tested in the presence of actors, but only C57L mice showed observational learning when tested in the absence of actors. The results with C57L like-sex pairs supported those with opposite-sex pairs (Experiment 6). Reactors' modelling, when tested either in the absence or in the presence of an actor, was further analysed in Experiments 7 and 8 respectively. It was found that the manipulation of certain variables markedly influenced reactors' modelling. These variables included: familiarity between conspecifics, aspects of approach and avoidance behaviour, food deprivation of subjects, type of screen separating subjects, and the extent of subjects' pre-test experience with the test apparatus.