Do autistic people see the 'average' of a group of objects?
Autistic adults, like neurotypical adults, automatically registered the average size of a group of circles even when they could not remember individual circles, challenging theories that autism blocks global processing.
Source
Perceptual Averaging in Individuals with Autism Spectrum Disorder
Study at a glance
- Design
- Human experiment — Within-subject psychophysics: mean-versus-member discrimination task (Exp 1) and mean-size adaptation aftereffect task (Exp 2), analysed separately in an autism group and a neurotypical control group
- N
- N=24 · 12 high-functioning autistic young adults and 12 neurotypical university students; the same participants completed both experiments
- Population
- Young adults (roughly 17-26 years) in Trento, Italy: autistic adults with IQ in the normal range and neurotypical students
- Outcome
- Accuracy in choosing the mean size versus a set member (Exp 1); proportion of 'right circle larger' responses after adapting to large vs small mean-size patches (Exp 2)
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
12 high-functioning autistic young adults and 12 neurotypical students did two short computer tasks. In Experiment 1 they saw 16 circles of different sizes for one second, then chose which of two test circles was the set's average size (mean task) or which had actually appeared (member task). In Experiment 2 they stared at two patches of circles, one with a larger average size, then judged which of two physically identical test circles looked bigger; a mean-size aftereffect would make the circle on the large-average side look smaller.
What they found
Both groups picked the mean size better than chance but were at chance on the member task, showing they represented the average without holding the individual sizes. Both groups also showed a clear adaptation aftereffect of similar size (19.3% for autistic participants and 20.67% for controls), even though averaging was irrelevant to the task. Controls showed the usual tendency to overestimate mean size, whereas the autistic group showed no sign of this bias, and their mean-versus-member advantage was about half the size of the controls'.
The limits
What it doesn't show
With only 12 people per group the study was designed to test whether averaging is present in autism, not to compare the groups, so the smaller effects and missing overestimation bias in the autistic group cannot be treated as real group differences. Participants were high-functioning autistic adults, so results may not apply across the spectrum, and IQ was not measured in controls. The member task may have been too hard for anyone to show an autistic advantage for detail.
Key terms
- Perceptual averaging (ensemble perception)
- The visual system's quick summary of a group of items, such as their average size, without encoding each item in detail.
- Adaptation aftereffect
- A perceptual bias after prolonged exposure to a stimulus, where later stimuli appear shifted in the opposite direction.
- Weak central coherence
- A theory that autistic people focus on details at the expense of integrating information into a global whole.
- Enhanced perceptual functioning
- A theory that autistic people have heightened access to low-level local detail without necessarily lacking global processing.
- Chance performance
- The accuracy expected from guessing alone, 50% in a two-choice task.
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 Experiment 1, how did both groups perform on the member task?
Common questions
Does this mean autistic and non-autistic people perceive averages identically?
Not necessarily. Both groups showed averaging, but the samples were too small to detect or rule out modest group differences, and the autistic group showed hints of a weaker mean advantage and no overestimation bias.
Why is Experiment 2 strong evidence that averaging is automatic?
Participants were only asked to judge individual circle sizes, yet their judgments were biased by the average size of the adapting patches, showing the average was encoded even though it was irrelevant and even unhelpful.
Which theories do these results challenge?
Theories predicting that autistic people fail to form global summaries unless instructed, such as weak central coherence and accounts of reduced use of prior information.
More on Autism