Trial-by-trial data analysis using computational models

Researchers have recently begun to integrate computational models into the analysis of neural and behavioural data, particularly in experiments on reward learning and decision making. This chapter aims to review and rationalize these methods. It exposes these tools as instances of broadly applicable statistical techniques, considers the questions they are suited to answer, provides a practical tutorial and tips for their effective use, and, finally, suggests some directions for extension or improvement. The techniques are illustrated with fits of simple models to simulated datasets. Throughout, the chapter flags interpretational and technical pitfalls of which authors, reviewers, and readers should be aware.