Hierarchical Bayesian estimation for cognitive models using Particle Metropolis within Gibbs (PMwG): A tutorial

Estimating quantitative cognitive models from data is a staple of modern psychological science, but can be difficult and inefficient. Particle Metropolis within Gibbs (PMwG) is a robust and efficient sampling algorithm that supports model estimation in a hierarchical Bayesian framework. This tutorial shows how cognitive modeling can proceed efficiently using pmwg, a new open-source package for the R language. We step through implementing the pmwg package with simple signal detection theory models, to more complex cognitive models in which two tasks are jointly modeled together. Through this process, we also address questions of model adequacy and model selection, which must be solved in order to answer meaningful psychological questions. PMwG, and the pmwg package, has the potential to move the field of psychology ahead in new and interesting directions, and to resolve questions that were once too hard to answer with previously available sampling methods.