Differential expectations produced by stimuli followed by nondifferential outcomes: A serial-compound view

Thomas Bryan DeMarse, Purdue University

Abstract

A series of five experiments investigated the possibility that animals may form differential outcome expectancies based on information provided by serial compounds of samples and comparisons. In Experiments 1 and 2, a transfer manipulation was arranged between the stimuli of two one-to-many differential outcome tasks where successful transfer required animals to form differential outcome expectancies based on common serial compound information hypothesized to occur in these tasks. In a second set of studies, the hypothesized compounds thought to produce the expectations responsible for transfer in Experiments 1 and 2 were removed in one group, and transfer was compared to a third nondifferential control group. Experiment 3 demonstrated that removal of the hypothesized serial compounds, and the information they provided, eliminated the successful transfer observed earlier, supporting the notion that animals were using this information to correctly choose among the comparisons during transfer. Experiments 4 and 5 further investigated the nature of these compounds. Experiment 4 demonstrated that the removal of one element of these compounds resulted in animals using other strategies to anticipate the outcomes in this task. Experiment 5 showed that novel stimuli associated with different outcomes are also able to successfully replace the sample stimuli of tasks in which these elements are removed, providing further support for the notion that animals had anticipated these outcomes in earlier experiments. Together, these experiments demonstrate that animals are able to anticipate different outcomes based on information provided by serial compounds of samples and comparisons even though these stimuli, by themselves, do not provide that information.

Degree

Ph.D.

Advisors

Urcuioli, Purdue University.

Subject Area

Psychology|Experiments

COinS