The role of working memory in visual statistical learning: a review

Visual statistical learning (VSL) is the ability to detect the regularities in the environment and exploit them to interact with our visual world in a more efficient way. Traditionally, this type of learning has been considered as automatic, incidental, inflexible, unconscious, and independent of availability of cognitive resources. The current manuscript reviews the role of working memory in VSL. In the studies that address this question, participants are asked to complete a visual search task where the presentation of the stimuli follows a statistical regularity. The search task is often combined with a concurrent secondary task that is intended to deplete participants’ working-memory resources. We discuss the evidence arising from studies that have addressed this question using popular statistical learning paradigms such as contextual cueing and location probability cueing. The contradictory results found in the (still scarce) literature on this topic lead us to conclude that the role of working memory in VSL is far from being completely understood.