Do autistic children look at a talker's mouth when they need to?
When listening to speech in noise, autistic children looked less at the talker's face and much less at the mouth than matched peers, instead lingering on uninformative parts of the face.
Source
Seeing to hear? Patterns of gaze to speaking faces in children with autism spectrum disorders
Study at a glance
- Design
- Case-control — Matched-group eye-tracking study: children watched syllables in auditory noise and silently (lipreading) plus a non-speech moving-shape control; gaze to face, mouth and non-focal regions analysed in 300 ms time bins.
- N
- N=20 · 20 verbal children: 10 with ASD and 10 typically developing controls matched on sex, age, cognitive and language scores (three ASD children were excluded for looking too little at the face).
- Population
- Monolingual English-speaking verbal school-age children with autism spectrum diagnoses, and matched typically developing peers.
- Outcome
- Percentage of looking time on the face, and within the face on the mouth region versus non-focal regions (cheeks, forehead, ears).
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Ten verbal children with autism and ten matched typically developing children watched a man say 'ma' or 'na', either with loud background noise or with the sound removed so they had to lipread, and reported what they perceived. A remote eye-tracker recorded where they looked across each 1.5-second syllable. As a control, they also watched figure-eight shapes that grew and shrank in time with the syllables, paired with tone-like sounds.
What they found
Autistic children spent less time on the face than controls in the noise task (about 61% vs 82% of on-screen time), with a smaller, marginal difference when lipreading. Of their face-looking time, they spent about half as much on the mouth in noise (26% vs 53%) and more on non-focal areas; typically developing children shifted to the mouth as the articulation unfolded, but autistic children did not in noise, though they showed a weaker version of this shift when lipreading. Gaze to the non-speech shapes did not differ between groups.
The limits
What it doesn't show
With only ten children per group, a wide age range and mixed diagnoses (autism, Asperger syndrome, PDD-NOS), the estimates are imprecise, as the authors caution. Three autistic children were excluded for looking too little at the face, so the sample may underrepresent those with the most atypical gaze. Stimuli were single syllables from one male speaker, not natural sentences, and the study cannot tell whether reduced mouth-looking causes the weaker audiovisual speech perception or whether processing differences also play a role.
Key terms
- Audiovisual speech perception
- Using both what you hear and the visible mouth movements of a speaker to identify speech, especially helpful in noise.
- Speechreading
- Identifying speech from the speaker's visible articulation alone, without sound (lipreading).
- Non-focal face regions
- Parts of the face such as cheeks, forehead and ears that carry little or no information about speech movements.
- Signal-to-noise ratio
- How loud the speech is relative to background noise; more negative values mean the speech is harder to hear.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
In the speech-in-noise task, autistic children looked relatively more at...
Common questions
Earlier studies said autistic people look more at the mouth. Why the difference here?
Those studies mostly used emotional or social judgements where the mouth was not the key information. Here the mouth carried the speech information, and autistic children looked at it less, suggesting gaze does not follow informational value in the usual way.
Why include the moving shapes?
To check whether the groups simply differ in tracking any dynamic audiovisual display. Because gaze to the shapes was similar, the differences seem specific to speaking faces.
More on Autism