Our perception of the world is clearly multisensory, but the weight that a single sensory modality adds to the final percept is sometimes difficult to determine. A simple way to investigate this issue has been to present conflicting sensory modalities, and to observe the sensory preference. For instance, the Illusory Flash Effect reported by Shams et al. (2002) shows that whenever a visual stimulus is presented together with one or more auditory beeps, adult participants tend to perceive multiple flashes. While the interaction between senses has been largely studied in the adult population, the development of non-linguistic sensory perception in school-aged children has remained underinvestigated. Recent studies have shown that children up to 4 years of age have an overall preference for auditory inputs (Robinson & Sloutsky, 2003). However, it remains to be ascertained how this auditory dominance may change during development.
Here we investigated audio-visual interactions in a developmental perspective, by testing the Illusory Flash Effect in school-aged children of 6, 9 and 11 years old. We adopted this paradigm (that avoids linguistic stimuli) under the assumption that auditory preference should be stronger in children of 6 years of age, and progressively decrease with age given that the overall sensory preference in adulthood typically favours the visual modality.
Visual stimuli consisted of a uniform yellow disk subtending 2° of visual field presented in the centre of the computer monitor. Auditory stimuli consisted of an 80 dB beep (4 kHz). Visual stimuli could be presented either alone (unisensory conditions: 1, 2 or 3 flashes) or combined with 1, 2 or 3 auditory beeps (multisensory conditions). Children sat approximately at 60 cm from the monitor and were asked to type on the keyboard the corresponding number of perceived flashes.
Results showed that all three groups of children presented an illusory effect. However, 6-year-old children made more illusory fusions (i.e., less flashes than actually presented) than the other two groups when 2 or 3 flashes were paired with 1 beep. Similarly, 6-years-old children made more illusory fissions (i.e., more flashes than actually presented) than the other two groups, when 1 flash was presented together with 2 or 3 beeps. Because all children recognised single flashes equally well when presented unimodally, the latter result particularly indicates that more illusory mistakes in 6-years-old cannot be reduced to counting difficulties.
Our results support the existence of different maturational patterns for auditory and visual modalities. Auditory preference decreases with age, switching towards an overall visual preference. This pattern may be a consequence of visual abilities improving with age, due to reading and speech skills learned at school, and that mainly rely on vision. Finally, although speculative, there may be attentional changes that occur with age, that automatically shift attention towards the preferred sensory modality.