Cognition Across Sensory Domains: Biasing Predictive Classification of Morphed Multimodal Stimuli through Sequence Learning

Abstract

The usefulness of temporal sequences for examining explicit learning has been well-established, but the precise implicit learning effects of sequential context on stimuli discriminability remain underexplored. To examine the efficacy of sequence learning on the modulation of stimulus expectation over a limited time range, we designed a behavioral psychophysics experiment in which the subject is taught binary associations using brief presentation of two prototypical stimuli, and then tested to what extent these associations persisted in morphs of the two stimuli when presented in a temporal sequence. The experiment was administered with sets of stimuli that were either wholly visual or auditory in composition. We hope to find that the categorization of morphed stimuli is biased by the suggested structure of the sequences, and that this effect generalizes across both the visual and auditory domains. These results would advance our understanding of cognitive flexibility as it relates to context-dependent behavior, and by extension would serve to advance practical knowledge of executive control in relation to sensory stimuli.