ReadMe.txt

% Anne Collins, UC Berkeley
% annecollins@berkeley.edu, 2024
% Code for paper "RL or not RL? Parsing the processes that support human
% reward-based learning."

Full code and data for the paper "RL or not RL? Parsing the processes that support human
 reward-based learning." By Anne GE Collins.
https://osf.io/preprints/psyarxiv/he3pm

This code is developed in MATLAB 2020B, and includes use of the “optim” toolbox and the “statistics” toolbox.

There are 3 folders corresponding to three independent parts of the paper. Please cd. Into each folder before running scripts.
- RLWM contains data and code for all RLWM data-sets modeling and analyses, including main figures 1 and 2, as well as multiple supplementary figures. See readme inside the folder.
- RLWMP contains data and code for RLWMP data-set modeling and analyses, including main figure 3 and multiple supplementary figures. See readme inside the folder.
- SimulationFig4 folder contains code for the simulation study shown in Fig. 4

Additionally, for convenience, the SPM package spm12 is included in this folder as a dependency for model comparison scripts. However, installation of this package may be needed to run the model comparison scripts. 
SPM is a package developed at UCL and can be downloaded freely from the source below.
https://www.fil.ion.ucl.ac.uk/spm/software/