Why Does Higher Working Memory Capacity Help You Learn?

Algorithms for approximate Bayesian inference, such as Monte Carlo methods, provide one source of models of how people may deal with uncertainty in spite of limited cognitive resources. Here, we model learning as a process of sequential sampling, or 'particle filtering', and suggest that an individual's working memory capacity (WMC) may be usefully modelled in terms of the number of samples, or 'particles', that are available for inference. The model qualitatively captures two distinct effects reported recently, namely that individuals with higher WMC are better able to (i) learn novel categories, and (ii) flexibly switch between different categorization strategies.