Hierarchical-Bayesian-Cognitive-Modeling


Link to outside resource:

How cognitive modeling can benefit from hierarchical Bayesian models, Journal of Mathematical Psychology - Link


Connection to what we’ve learned:

The article described methods of modeling of cognitive processes in the realm of psychology through hierarchical Bayesian models. This paper connected to what we covered in class through Bayes Theorem, and probability distributions. The authors expanded on this groundwork and argued for the use of hierarchical models as opposed to non hierarchical ones which are just a function with parameters creating observed data. The author believed that by creating a hierarchy of models, such as by modeling the creation of observed function parameters(expanding to become ) and through this process getting more insights into the processes underlying cognition.

This articles connects back to what we’ve learned by continuing on from the realm of hypothesis testing to attempt to create falsifiable models of behavior through hierarchical modeling, highlight a number of domains where it could be used. The author highlights models in developmental learning, showing that behavior and stage of development together could be used to model more aspects of the data, by using Bayesian inference to combine the information through their hierarchical model to understand how they create developmental learning. It’s quite interesting seeing how statistics can be used further than testing statistical hypothesis, to trying to understand underlying probabilistic processes.