Discrepancy and nondiscrepancy methods of assessing visual-auditory interaction.

Abstract

Much of our information about interaction among the sensory modalities has come from studies in which discrepancy between the modalities has been imposed experimentally. The question arises whether the conclusions generated from discrepancy research are affected by the use of the nonnatural experimental situation: Would the same conclusions be reached if discrepancy were not used? In addition to using the index of interaction derived from discrepancy, as in prior research, the present study assessed two nonexperimentally imposed indices to explore the interaction between vision and audition in a spatial localization task. The additional indices, or “tags,” were standard deviations (SDs) and confidence ratings (CRs) of localization responses. The SD tag showed a pattern of interaction between vision and audition that was identical to that found with spatial discrepancy. The CR tag produced a similar pattern but did not show the same regular variation with the independent variable. Both SD and CR tags showed identical patterns of results when an experimentally imposed discrepancy was and was not used, suggesting that the use of moderate degrees of experimental discrepancy does not produce misleading conclusions about intersensory interaction.