Effect of the Visual Presentation of a Craniofacial Syndrome on Speech Intelligibility in Noise.

ABSTRACT

OBJECTIVE: Research has argued that a speaker's facial appearance can result in an "intelligibility cost" for the listener. The study investigated whether such an intelligibility cost exists for a visible repaired cleft lip and nasal asymmetry. SETTING: University department. PARTICIPANTS: Eight typical speakers provided speech samples. Twenty-eight naive listeners participated in a speech in noise experiment. INTERVENTIONS: Listeners transcribed sentences in noise that were paired with faces of individuals with repaired cleft lip and nasal asymmetry or typical faces. They also rated speaker intelligibility and answered a questionnaire about their previous knowledge about cleft lip and palate. MAIN OUTCOME MEASURES: Percentage of words transcribed correctly and intelligibility ratings, compared by experimental condition (photo of typical face or face with repaired cleft lip and nasal asymmetry) and speaker gender. RESULTS: There were no statistically significant differences between speech stimuli that were presented with faces with repaired cleft lip and nasal asymmetry or typical faces. The percentage of words transcribed correctly by the listeners was lower for female speakers (F = 12.7, df = 1; P < .01). Speech intelligibility of female speakers was rated more poorly (F = 10.5, df = 1; P < .01). CONCLUSIONS: Presence of a repaired cleft lip and nasal asymmetry did not result in an intelligibility cost for naive listeners. Future research should investigate possible effects of facial motion or previous knowledge.