Reconsidering the McGurk effect

The McGurk effect was studied on 3 groups of subjects differing in age: pre-schoolers (mean: 6 years), elementary-school 4-graders (mean: 11 years) and adults (mean: 24 years). The stimuli included audio-visual combinations of syllables /pa/, /ta/, /ka/, /ba/, /da/, /ga/. The combination of auditorily presented (A) bilabials with visually presented (V) velars, was reported by some subjects as a third (dental, e.g. /da/) plosive, and in the reversed-modality combination as a bilabial+velar response (e.g. /bga/), manifesting the McGurk effect. In all groups, dental–velar pairs and Adentals combined with Vbilabials elicited close to 100% responses corresponding to the A stimuli. Vdentals combined with Abilabials elicited an unexpectedly high proportion of responses corresponding to the visual stimuli. The McGurk effect seems less robust than commonly reported and exhibits individual variation. Group differences were inconsistent and the effect strength could not be attributed to age.