Abstract: The existing research paradigms of multisensory integration usually involve the random presentation of stimuli of unisensory modalities and bi-sensory modality. Such kind of paradigms, however, exists the modality shift effect which could lead to inaccurate measurement of multisensory integration. Therefore, it is necessary to exam the modality shift effect and separate it from the tests in order to analyze the results of multisensory integration.
Experiment one examined whether modality shift effect affected the audiovisual integration in the previous paradigms of multisensory integration. The visual, auditory, and audiovisual stimuli were randomly presented. Subjects were suggested to respond to those stimuli as quickly as they could. Results showed that the modality types of pre-stimuli affected the response to the post-stimuli. The race modal is invalid only when the pre-stimuli were from the audiovisual or visual modality.
Experiment two manipulated the intensity of pre-and-post stimuli so as to examine whether modality shift effect was caused by the suppression of reaction during the switch of modalities or by the priming on the same modality. The subjects were asked to response when detected visual or auditory stimulus. Through the analysis of the subjects’ reactions, we found that the intensity of pre-stimuli affected the response to the post-stimuli.
Therefore, we concluded that among the paradigms of multisensory integration, two factors had effects on the measurement of multisensory integration: First, dynamic adjusting on distribution of attention affected by the pre-stimuli; second, dynamic adjusting on the alertness of present reaction caused by the intensity of pre-stimuli. To get the pure results in multisensory measuring, these two influence factors should be eliminated in the measurement of multisensory integration.

Key words: multisensory integration, modality shift, race modal, coactivation model