The data are reported on in the project " A computational Model of Non-optimal Suspiciousness in the Minnesota Trust Game" by Rebecca Kazinka, Iris Vilares, & Angus W. MacDonald (University of Minnesota). Analyses were completed on Matlab 2019. We divided the data into two different folders, behavior and scripts. Main analysis scripts are found in the ‘scripts’ folder. Analyses for the paper were conducted in the file GenerateGraphs.m. This relies on the data located in the ‘behavior’ folder, which is currently unavailable due to IRB reasons. There are several mat files in the behavior folder labeled for each sample and analysis type. In the scripts folder, you can also find MTGWrapper_Sample1.m and MTGWrapper_Sample2.m, which run model-agnostic analyses on the original data to get summary results (found in the behavior folder – Sample1_OriginalData.csv & Sample2_OriginalData.csv). The summaries are saved in Sample1_output.csv and Sample2_output.csv. Threshold analyses rely on the script fitHeavisideThresholdMTG.m. organizeMTGDatafile_Sample1.m was used to organize the original Sample1 data into a format for modeling. Modeling was completed using the scripts fitModelsToMTGperSubject_fmincon_Sample1.m, fitModelsToMTGperSubject_fmincon_Sample2.m, and fitModelsToMTGperSubjectperDecisionAgent_fmincon_Sample1.m. These scripts would implement the model fitting based on the file name input from the model_scripts folder (inside the scripts folder). This would require entering the text after the ‘getNLL’ script for whichever model you wanted to implement. One would also need to change the parameters included to match those tested in the model. The getNLL script calls on the corresponding script to apply the model and calculate the negative log likelihood. Finally, the simulated data is created using the simModelsMTG_Sample1.m or simModelsMTG_Sample2.m analysis script. You will need to save the parameters from the model fit (fittedRealValParamPerSubject variable) and import it into the script. It will generate new behavioral data based on the parameters provided from the model fitting for the actual subjects, which can then be used to create a graph of the data. This script relies on the fitSimulation scripts in the model_scripts folder. ModelTestingSummary.csv shows the results for all attempted model fits seen in the supplemental materials. You will also need the following non-standard functions, which can be found as part of Matlab's AddOn packages. - mengz - ColorBrewer - LOESS regression smoothing