Reduced Context Updating but Intact Visual Priors in Autism
A general consensus persists that sensory-perceptual differences in autism, such as hypersensitivities to light or sound, result from an overreliance on new (rather than prior) sensory observations. However, conflicting Bayesian accounts of autism remain unresolved as to whether such alterations are...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Ubiquity Press
2021-12-01
|
| Series: | Computational Psychiatry |
| Subjects: | |
| Online Access: | https://cpsyjournal.org/articles/69 |
| _version_ | 1828420220362948608 |
|---|---|
| author | R. Randeniya I. Vilares J. B. Mattingley M. I. Garrido |
| author_facet | R. Randeniya I. Vilares J. B. Mattingley M. I. Garrido |
| author_sort | R. Randeniya |
| collection | DOAJ |
| description | A general consensus persists that sensory-perceptual differences in autism, such as hypersensitivities to light or sound, result from an overreliance on new (rather than prior) sensory observations. However, conflicting Bayesian accounts of autism remain unresolved as to whether such alterations are caused by more precise sensory observations (precise likelihood model) or by forming a less precise model of the sensory context (hypo-priors model). We used a decision-under-uncertainty paradigm that manipulated uncertainty in both likelihoods and priors. Contrary to model predictions we found no differences in reliance on likelihood in autistic group (AS) compared to neurotypicals (NT) and found no differences in subjective prior variance between groups. However, we found reduced context adjustment in the AS group compared to NT. Further, the AS group showed heightened variability in their relative weighting of sensory information (vs. prior) on a trial-by-trial basis. When participants were aligned on a continuum of autistic traits, we found no associations with likelihood reliance or prior variance but found an increase in likelihood precision with autistic traits. These findings together provide empirical evidence for intact priors, precise likelihood, reduced context updating and heightened variability during sensory learning in autism. |
| first_indexed | 2024-12-10T15:08:11Z |
| format | Article |
| id | doaj.art-a7c6f0bffdf44936900b424074cd5e76 |
| institution | Directory Open Access Journal |
| issn | 2379-6227 |
| language | English |
| last_indexed | 2024-12-10T15:08:11Z |
| publishDate | 2021-12-01 |
| publisher | Ubiquity Press |
| record_format | Article |
| series | Computational Psychiatry |
| spelling | doaj.art-a7c6f0bffdf44936900b424074cd5e762022-12-22T01:44:00ZengUbiquity PressComputational Psychiatry2379-62272021-12-015110.5334/cpsy.6963Reduced Context Updating but Intact Visual Priors in AutismR. Randeniya0I. Vilares1J. B. Mattingley2M. I. Garrido3Queensland Brain Institute, The University of QueenslandDepartment of Psychology, University of MinnesotaQueensland Brain Institute, The University of Queensland; School of Psychology, The University of Queensland, AU; Canadian Institute for Advanced Research (CIFAR), CA; Australian Research Council Centre of Excellence for Integrative Brain FunctionMelbourne School of Psychological Sciences, University of Melbourne; Australian Research Council Centre of Excellence for Integrative Brain FunctionA general consensus persists that sensory-perceptual differences in autism, such as hypersensitivities to light or sound, result from an overreliance on new (rather than prior) sensory observations. However, conflicting Bayesian accounts of autism remain unresolved as to whether such alterations are caused by more precise sensory observations (precise likelihood model) or by forming a less precise model of the sensory context (hypo-priors model). We used a decision-under-uncertainty paradigm that manipulated uncertainty in both likelihoods and priors. Contrary to model predictions we found no differences in reliance on likelihood in autistic group (AS) compared to neurotypicals (NT) and found no differences in subjective prior variance between groups. However, we found reduced context adjustment in the AS group compared to NT. Further, the AS group showed heightened variability in their relative weighting of sensory information (vs. prior) on a trial-by-trial basis. When participants were aligned on a continuum of autistic traits, we found no associations with likelihood reliance or prior variance but found an increase in likelihood precision with autistic traits. These findings together provide empirical evidence for intact priors, precise likelihood, reduced context updating and heightened variability during sensory learning in autism.https://cpsyjournal.org/articles/69asdvisualsensitivitiesautism traitsbayesianinsistence on samenessrestrictive repetitive behavioursrrbpriorlikelihood |
| spellingShingle | R. Randeniya I. Vilares J. B. Mattingley M. I. Garrido Reduced Context Updating but Intact Visual Priors in Autism Computational Psychiatry asd visual sensitivities autism traits bayesian insistence on sameness restrictive repetitive behaviours rrb prior likelihood |
| title | Reduced Context Updating but Intact Visual Priors in Autism |
| title_full | Reduced Context Updating but Intact Visual Priors in Autism |
| title_fullStr | Reduced Context Updating but Intact Visual Priors in Autism |
| title_full_unstemmed | Reduced Context Updating but Intact Visual Priors in Autism |
| title_short | Reduced Context Updating but Intact Visual Priors in Autism |
| title_sort | reduced context updating but intact visual priors in autism |
| topic | asd visual sensitivities autism traits bayesian insistence on sameness restrictive repetitive behaviours rrb prior likelihood |
| url | https://cpsyjournal.org/articles/69 |
| work_keys_str_mv | AT rrandeniya reducedcontextupdatingbutintactvisualpriorsinautism AT ivilares reducedcontextupdatingbutintactvisualpriorsinautism AT jbmattingley reducedcontextupdatingbutintactvisualpriorsinautism AT migarrido reducedcontextupdatingbutintactvisualpriorsinautism |