**Materials** Task can be viewed here: https://app.gorilla.sc/openmaterials/802718 Supplementary materials are included with files below ---------------------------------------------- **Data** Open zipped folder OSF In the folder Analysis you will find *Aggregate data* **rt.csv** and **acc.csv** files containing each participants average response time and accuracy data in the congruent and incongruent conditions. **param_est.csv** file containing the aggregated output from the model fitting. Boundary Separation (A), Non-Decision Time (ter), perceptual processing (p), spotlight rate (rd), spotlight standard deviation (sda), he likelihood ratio chi-square statistic (G2), Binned Bayesian Information Criterion (bBIC), group, Spotlight (interference time, Sda/rd). *R Script* **analysis_osf.R** is the r script used to clean the raw data, analyse and plot response time data and analyse and plot the estimated model parameters. To run the script set the wd to the folder where you have unpackaged the zip folder In the Autistic and NT folders you will find: **rawdata.csv** this is the raw flanker task output from Gorilla. The participant responses are on rows where ZoneType == response_keyboard. The correct column indicates if the response was correct = 1 or incorrect = 0; Reaction Time gives the Reaction Time, Type indicates if the trial was congruent or incongruent. For the autistic group Participant.Public.ID gives the participant number - note the values do not increase linearly as participants were assigned a number if they consented to be contacted about the study, not all contacted participants took part. For the non-autistic group participant id is Participant.Private.ID as assigned by prolific. **RAADS_14.csv** this is the raw RAADS-14 output from Gorilla. The final columns give the question number (1 -14 + an attention check) and the response gives the participant response **CFQ.csv** this is the raw CFQ output from Gorilla. The CFQ was developed from a clone from the Gorilla online repository (https://app.gorilla.sc/openmaterials/50646). The responses are recorded where Question.Key == response-n-quantised, response is recorded in the Response column In the Model Fitting folder you will find: Everything used for fitting data to the Shrinking Spotlight Model. These model fitting was conducted using the High Performance Computing cluster at the University of Sheffield. The reason for this is the fitting of data to SSP using flankR is extremely slow. Running the batches of 150 participants on the HPC took ~9 days **my_job** is a batch script for running the RScript for the Autistic Participants **my_job_2** is the batch script for non-autistic participants **model_fitting_ind** this R script estimates each individual participants best fitting parameters as described in the manuscript. The script reads in the numbered **n.csv** files which contain the individual participant accuracy and response time data formatted to be read by flankR (output from **analysis_osf.R**) For each participant **model_fitting_ind** outputs: a .csv containing the estimated parameters for that individual (this is what is aggregated in **param_est.csv**). a .csv containing 1000 simulated trials based on the best fitting parameters. The data from Model Fitting is also separated out by group in the Autistic and NT folders organised into Model Fits 50 (giving the individual model fits) and Simulated 50 (giving the similuated data) Note - following consultation with autistic consultants through the design of the study we decided not to make ICAR-16 nor demographic data available with the study.