Abstract
The ability to infer the goals and intentions of others is crucial for social interactions, and such social capabilities are broadly distributed across individuals. Autism-like traits (that is, traits associated with autism spectrum disorder (ASD)) have been associated with reduced social inference, yet the underlying computational principles and social cognitive processes are not well characterized. Here we tackle this problem by investigating inference during social learning through computational modeling in two large cross-sectional samples of adult participants from the general population (N1 = 943, N2 = 352). Autism-like traits were extracted and isolated from other associated symptom dimensions through a factor analysis of the Social Responsiveness Scale. Participants completed an observational learning task to quantify the tradeoff between two social learning strategies: imitation (repeating the observed partner’s most recent action) and emulation (inferring the observed partner’s goal). Autism-like traits were associated with reduced observational learning specifically through reduced emulation (but not imitation), revealing difficulties in social goal inference (Pearson’s r = −0.124, P < 0.001). This association held, even when controlling for other model parameters (for example, decision noise, heuristics, F1,925 = 15.352, P < 0.001), and was specifically related to social difficulties in autism-like traits (F1,916 = 33.169, P < 0.001) but not social anxiety traits (F1,916 = 0.005, P = 0.945). The findings, replicated in an additional sample, provide a powerfully specific mechanistic hypothesis for social learning challenges in ASD, employing a computational psychiatry approach that could be applied to other disorders.
This is a preview of subscription content, access via your institution
Access options
Subscribe to this journal
Receive 12 digital issues and online access to articles
$79.00 per year
only $6.58 per issue
Buy this article
- Purchase on SpringerLink
- Instant access to full article PDF
Prices may be subject to local taxes which are calculated during checkout
Similar content being viewed by others
Data availability
All raw data and curated data spreadsheets are publicly available at https://osf.io/j5npu/ and https://github.com/wuqy052/ASD_ObsLearn.
Code availability
All codes for the experiment, as well as analysis scripts, can be found online at https://github.com/wuqy052/ASD_ObsLearn.
References
Lundström, S. et al. Autism spectrum disorders and autistic like traits: similar etiology in the extreme end and the normal variation. Arch. Gen. Psychiatry 69, 46–52 (2012).
Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 2013); https://doi.org/10.1176/appi.books.9780890425596
Ruzich, E. et al. Measuring autistic traits in the general population: a systematic review of the Autism-Spectrum Quotient (AQ) in a nonclinical population sample of 6,900 typical adult males and females. Mol. Autism 6, 2 (2015).
Leadbitter, K., Buckle, K. L., Ellis, C. & Dekker, M. Autistic self-advocacy and the neurodiversity movement: implications for autism early intervention research and practice. Front. Psychol. 12, 635690 (2021).
Russell, G. in Autistic Community and the Neurodiversity Movement: Stories from the Frontline (ed. Kapp, S. K.) 287–303 (Springer, 2020); https://doi.org/10.1007/978-981-13-8437-0_21
Henco, L. & Schilbach, L. Studying social inferences in and across social brains. Biol. Psychiatry Cogn. Neurosci. Neuroimaging 6, 760–761 (2021).
Rusch, T., Steixner-Kumar, S., Doshi, P., Spezio, M. & Gläscher, J. Theory of mind and decision science: towards a typology of tasks and computational models. Neuropsychologia 146, 107488 (2020).
Foti, F. et al. Observational learning in low-functioning children with autism spectrum disorders: a behavioral and neuroimaging study. Front. Psychol. 9, 2737 (2019).
Rosenthal, I. A., Hutcherson, C. A., Adolphs, R. & Stanley, D. A. Deconstructing theory-of-mind impairment in high-functioning adults with autism. Curr. Biol. 29, 513–519 (2019).
Hudson, M., Nijboer, T. C. W. & Jellema, T. Implicit social learning in relation to autistic-like traits. J. Autism Dev. Disord. 42, 2534–2545 (2012).
Charpentier, C. J., Iigaya, K. & O’Doherty, J. P. A neuro-computational account of arbitration between choice imitation and goal emulation during human observational learning. Neuron 106, 687–699.e7 (2020).
Charpentier, C. J. & O’Doherty, J. P. in The Neural Basis of Mentalizing (eds Gilead, M. & Ochsner, K. N.) 489–501 (Springer, 2021); https://doi.org/10.1007/978-3-030-51890-5_24
Call, J. & Carpenter, M. in Imitation in Animals and Artifacts (eds Dautenhahn, K. & Nehaniv, C. L.) 211–228 (Boston Review, 2002); https://doi.org/10.7551/mitpress/3676.001.0001
Baker, C. L., Jara-Ettinger, J., Saxe, R. & Tenenbaum, J. B. Rational quantitative attribution of beliefs, desires and percepts in human mentalizing. Nat. Hum. Behav. 1, 0064 (2017).
Boorman, E. D., O’Doherty, J. P., Adolphs, R. & Rangel, A. The behavioral and neural mechanisms underlying the tracking of expertise. Neuron 80, 1558–1571 (2013).
Collette, S., Pauli, W. M., Bossaerts, P. & O’Doherty, J. Neural computations underlying inverse reinforcement learning in the human brain. eLife 6, e29718 (2017).
Heyes, C. & Saggerson, A. Testing for imitative and nonimitative social learning in the budgerigar using a two-object/two-action test. Anim. Behav. 64, 851–859 (2002).
Nielsen, M. Copying actions and copying outcomes: social learning through the second year. Dev. Psychol. 42, 555–565 (2006).
Burke, C. J., Tobler, P. N., Baddeley, M. & Schultz, W. Neural mechanisms of observational learning. Proc. Natl Acad. Sci. USA 107, 14431–14436 (2010).
Want, S. C. & Harris, P. L. How do children ape? Applying concepts from the study of non-human primates to the developmental study of ‘imitation’ in children. Dev. Sci. 5, 1–13 (2002).
Horner, V. & Whiten, A. Causal knowledge and imitation/emulation switching in chimpanzees (Pan troglodytes) and children (Homo sapiens). Anim. Cogn. 8, 164–181 (2005).
Robic, S. et al. Decision-making in a changing world: a study in autism spectrum disorders. J. Autism Dev. Disord. 45, 1603–1613 (2015).
Sevgi, M., Diaconescu, A. O., Henco, L., Tittgemeyer, M. & Schilbach, L. Social Bayes: using Bayesian modeling to study autistic trait-related differences in social cognition. Biol. Psychiatry 87, 185–193 (2020).
Rosenblau, G., Frolichs, K. & Korn, C. W. A neuro-computational social learning framework to facilitate transdiagnostic classification and treatment across psychiatric disorders. Neurosci. Biobehav. Rev. 149, 105181 (2023).
Huys, Q. J. M., Browning, M., Paulus, M. P. & Frank, M. J. Advances in the computational understanding of mental illness. Neuropsychopharmacol. 46, 3–19 (2021).
Jacob, S. et al. Neurodevelopmental heterogeneity and computational approaches for understanding autism. Transl. Psychiatry 9, 63 (2019).
Wise, T., Robinson, O. J. & Gillan, C. M. Identifying transdiagnostic mechanisms in mental health using computational factor modeling. Biol. Psychiatry 93, 690–703 (2023).
Yoshida, W. et al. Cooperation and heterogeneity of the autistic mind. J. Neurosci. 30, 8815–8818 (2010).
Rosenblau, G., Korn, C. W., Dutton, A., Lee, D. & Pelphrey, K. A. Neurocognitive mechanisms of social inferences in typical and autistic adolescents. Biol. Psychiatry Cogn. Neurosci. Neuroimag. 6, 782–791 (2021).
Goris, J. et al. Autistic traits are related to worse performance in a volatile reward learning task despite adaptive learning rates. Autism 25, 440–451 (2021).
Crawley, D. et al. Modeling flexible behavior in childhood to adulthood shows age-dependent learning mechanisms and less optimal learning in autism in each age group. PLoS Biol. 18, e3000908 (2020).
Lawson, R. P., Mathys, C. & Rees, G. Adults with autism overestimate the volatility of the sensory environment. Nat. Neurosci. 20, 1293–1299 (2017).
Patzelt, E. H., Hartley, C. A. & Gershman, S. J. Computational phenotyping: using models to understand individual differences in personality, development and mental illness. Personal Neurosci. 1, e18 (2018).
Constantino, J. N. & Gruber, C. P. Social Responsiveness Scale: SRS-2 (Western Psychological Services, 2012).
Hu, L. & Bentler, P. M. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct. Equ. Modeling 6, 1–55 (1999).
Bejerot, S., Eriksson, J. M. & Mörtberg, E. Social anxiety in adult autism spectrum disorder. Psychiatry Res. 220, 705–707 (2014).
Baron-Cohen, S. in International Review of Research in Mental Retardation Vol. 23 (ed. Glidden, L. M.) 169–184 (Academic Press, 2000).
Frith, C. & Frith, U. Theory of mind. Curr. Biol. 15, R644–R645 (2005).
Schilbach, L. Towards a second-person neuropsychiatry. Phil. Trans. R. Soc. B 371, 20150081 (2016).
Schuwerk, T., Jarvers, I., Vuori, M. & Sodian, B. Implicit mentalizing persists beyond early childhood and is profoundly impaired in children with autism spectrum condition. Front. Psychol. 7, 1696 (2016).
White, S. W., Schry, A. R. & Kreiser, N. L. in Handbook of Autism and Anxiety (eds Davis III, T. E. et al.) 121–136 (Springer, 2014); https://doi.org/10.1007/978-3-319-06796-4_9
Sandhu, T. R., Xiao, B. & Lawson, R. P. Transdiagnostic computations of uncertainty: towards a new lens on intolerance of uncertainty. Neurosci. Biobehav. Rev. 148, 105123 (2023).
Sapey-Triomphe, L.-A., Weilnhammer, V. A. & Wagemans, J. Associative learning under uncertainty in adults with autism: intact learning of the cue-outcome contingency, but slower updating of priors. Autism 26, 1216–1228 (2022).
Yu, A. J. & Dayan, P. Uncertainty, neuromodulation and attention. Neuron 46, 681–692 (2005).
Payzan-LeNestour, E. & Bossaerts, P. Risk, unexpected uncertainty and estimation uncertainty: Bayesian learning in unstable settings. PLoS Comput. Biol. 7, e1001048 (2011).
Raven, J. & Raven, J. in Handbook of Nonverbal Assessment (ed. McCallum, R. S.) 223–237 (Kluwer Academic/Plenum Publishers, 2003); https://doi.org/10.1007/978-1-4615-0153-4_11
Jones, G. & Macken, B. Questioning short-term memory and its measurement: why digit span measures long-term associative learning. Cognition 144, 1–13 (2015).
Liebowitz, M. R. Social phobia. Mod. Probl. Pharmacopsychiatry 22, 141–173 (1987).
Piray, P., Dezfouli, A., Heskes, T., Frank, M. J. & Daw, N. D. Hierarchical Bayesian inference for concurrent model fitting and comparison for group studies. PLoS Comput. Biol. 15, e1007043 (2019).
Lloyd, S. Least squares quantization in PCM. IEEE Trans. Inf. Theory 28, 129–137 (1982).
Rousseeuw, P. J. Silhouettes: a graphical aid to the interpretation and validation of cluster analysis. J. Comput. Appl. Math. 20, 53–65 (1987).
Horn, J. L. A rationale and test for the number of factors in factor analysis. Psychometrika 30, 179–185 (1965).
Revelle, W. Psych: Procedures for Psychological, Psychometric and Personality Research (Northwestern University, 2021).
R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2021).
Rosseel, Y. lavaan: an R package for structural equation modeling. J. Stat. Softw. 48, 1–36 (2012).
Otto, A. R., Raio, C. M., Chiang, A., Phelps, E. A. & Daw, N. D. Working-memory capacity protects model-based learning from stress. Proc. Natl Acad. Sci. USA 110, 20941–20946 (2013).
Schad, D. J. et al. Processing speed enhances model-based over model-free reinforcement learning in the presence of high working memory functioning. Front. Psychol. 5, 1450 (2014).
Acknowledgements
This work was supported by funding from the NIMH R21MH120805 to J.P.O. and K99MH123669 to C.J.C., and the NIMH Caltech Conte Center on the Neurobiology of Social Decision-Making (P50MH094258 to J.P.O.) as well as by Caltech’s T&C Chen Center for Social and Decision Neuroscience (to J.P.O.). Q.W. was funded in part by a grant from the Simons Foundation Autism Initiative to R. Adolphs, and by Caltech’s T&C Chen Center for Social and Decision Neuroscience. We thank R. Adolphs for his helpful comments on the paper.
Author information
Authors and Affiliations
Contributions
Study conceptualization was provided by Q.W., R.T., J.D.F., J.C., J.P.O. and C.J.C. Task design was implemented by S.O., J.C., J.P.O. and C.J.C. Data collection and curation were carried out by Q.W., S.O., J.C. and C.J.C. The conceptualization of the computational models was provided by Q.W., J.C., J.P.O. and C.J.C. Data analyses were performed by Q.W., S.O. and C.J.C. The original draft was written by Q.W. and C.J.C. Review and editing were carried out by Q.W., S.O., R.T., J.D.F., J.C., J.P.O. and C.J.C. J.P.O. and C.J.C. jointly supervised the work. Funding acquisition was carried out by J.D.F., J.C., J.P.O. and C.J.C.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Mental Health thanks Leonhard Schilbach and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data
Extended Data Fig. 1 Participants inclusion pipeline.
Flow diagram showing the number of participants included/excluded at each recruitment and analysis step. *for all the exclusion criteria, check Supplementary Method 2.
Extended Data Fig. 2 Visualization of the clusters in the model-fitting feature space.
The distribution of individual model fitting diagnostics across each sample, shown as scatter plots for every pair of model-fitting diagnostics used as clustering dimensions. Colors represent four strategy groups. R2 Imitation: pseudo R2 of the imitation model, R2 NonLearn: pseudo R2 of the non-learning model, R2 Emulation: pseudo R2 of the emulation model, R2 Fix-Single: difference between the pseudo R2 of the fixed mixture model and the better single strategy model (imitation or emulation), R2 Dyn-Fix: difference between the pseudo R2 of the dynamic arbitration model and the fixed mixture model.
Extended Data Fig. 3 Exploratory factor analysis of SRS item data.
(a) Parallel analysis on the discovery sample. The optimal number of factors is 8, determined by the maximum number before the scree plots from the actual data and resampled simulated data intersect. (b) Cumulative variance explained by each factor. Factors were sorted in descending order of their explained variance.
Extended Data Fig. 4 Loadings of SRS Items on factors.
Loading strength of each of the 65 SRS items on each of the 8 factors. Orange indicates positive loadings, and blue indicates negative loadings. The color intensity is proportional to the loading strength. The number before each statement corresponds to the item index from the SRS-2 ASR questionnaire.
Extended Data Fig. 5 Correlations between arbitration diagnostics and autism-like traits.
(a) Correlation between arbitration propensity and factor 1 score. Discovery sample: N = 943, Pearson’s r = −0.016, p = 0.629, two-tailed; replication sample: N = 352, Pearson’s r = −0.012, p = 0.819, two-tailed. (b) Correlation between uncertainty weight parameter η and factor 1 score. Discovery sample: N = 943, Pearson’s r = −0.031, p = 0.343, two-tailed; replication sample: N = 352, Pearson’s r = −0.018, p = 0.733, two-tailed. Error bands represent the 95% confidence interval.
Extended Data Fig. 6 Comparison between AIC-based clustering and unsupervised clustering approaches.
The ‘Lowest AIC Approach’ classifies individuals using a fixed boundary. The ‘Unsupervised Clustering’ represents our current data-driven subject clustering approach, which does not have a hard boundary between groups, but clusters individuals based on the cosine distance of dimension values in a high-dimensional space. The top right (green) cluster is a separate subject group (Group 3) in the latter approach but is split into two other groups (Group 1, Group 2) in the former approach.
Supplementary information
Supplementary Information
Supplementary Methods 1–7, Results 1–6, Discussions 1–3, Figs. 1–9, Tables 1 and 2, and References.
Supplementary Table 2
Table S2. SRS factor analysis results. Loadings of each item on their corresponding factors, ranked by the magnitude of the loadings within each factor. An item is considered as part of a factor if its loading on this particular factor is the largest among all factors, and the loading is more than 0.3. MOT, social motivation; AWR, social awareness; RRB, restricted interests and repetitive behaviors; COG, social cognition; COM, social communication.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Wu, Q., Oh, S., Tadayonnejad, R. et al. Individual differences in autism-like traits are associated with reduced goal emulation in a computational model of observational learning. Nat. Mental Health 2, 1032–1044 (2024). https://doi.org/10.1038/s44220-024-00287-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1038/s44220-024-00287-1
