Abstract:

The sound-induced flash illusion (SIFI) is an auditory-dominant phenomenon in multisensory integration in which the perception of brief visual flashes may be qualitatively altered by concurrent brief sounds. It has been well documented how attention modulates this illusion. It remains unknown, however, how modal-based endogenous attention interferes with the SIFI. In the present study, we adopted the classical paradigm of the SIFI and directed the endogenous attention to the visual modal or the auditory modal to investigate the effect of modal-based endogenous attention on SIFI.

Experiment 1 asked the participants to conduct the classical task of the SIFI as the baseline. In experiment 2, the ratio of the visual modal, the auditory modal, and the audio-visual modal was 3:1:1; that is, the endogenous attention was directed to the visual modal. In experiment 3, the ratio of the visual modal, auditory modal and audio-visual modal was 1:3:1; that is, the endogenous attention was directed to the auditory modal. In the present study, we asked the participants to judge the number of flashes, and we focused mainly on the occurrence of fission illusions (when a single visual flash is accompanied by two auditory bleeps and is perceived incorrectly as two flashes) and fusion illusions (when two visual flashes are accompanied by a single bleep and are perceived incorrectly as a single flash). The hypothesis was that directing endogenous attention to the visual or auditory modal could affect the auditory dominance.

A repeated measures ANOVA was analyzed for fission and fusion illusions separately, and the results showed that there were significant differences among the three experiments, F (1, 2) = 5.11, p < 0.01, η ²= 0.23. The occurrence of fission illusions was decreased when endogenous attention was directed to the visual modal (50%) and was increased when endogenous attention was directed to the auditory modal (71%). However, regardless of whether endogenous attention was directed to the auditory or visual modal, fusion illusion was not affected by attention (ps > 0.05). In addition, we also pooled the response counts across all of the participants and for both fission and fusion stimuli. The results showed that the odds ratio for fission was greater than that for fusion, and there was a difference among the three experiments for fission (17.5 vs. 9 vs. 21.95), and no effect for fusion (4.95 vs. 3.15 vs. 4.13).

The results indicated that modal-based endogenous attention can affect the occurrence of fission illusions but does not affect the occurrence of fusion illusions. The present study provides insight into the top-down factors that can modulate sound-induced flash illusions.

Key words: endogenous attention, auditory dominance, sound-induced flash illusion, visual and auditory modals