Contextual influences on visual statistical learning

Abstract

Visual statistical learning describes the unintentional extraction of statistical regularities from visual environments across time or space, and is typically studied using novel stimuli (e.g., symbols unfamiliar to participants). Additionally, familiarization procedures in experiments that have explored visual statistical learning tend to be passive or require only basic vigilance from participants. The natural visual world, however, is rich with a variety of complex visual stimuli, and we experience that world in the presence of goal-driven behavior including overt learning of other kinds. The present dissertation examines how visual statistical learning responds to such contextual factors. Chapter 2 finds that visual statistical learning is sometimes influenced by natural and artificial categories in the presence (and absence) of explicit learning about artificial categories. Chapter 3 examines the impacts of categories on visual statistical learning in the presence of systematic visual similarity manipulations, and also considers how visual similarity might be modulated by statistical learning. Chapter 4 focuses on how different familiarization tasks may influence the behavioral and neural correlates of visual statistical learning using brain imaging (fMRI). Together, the results from these experiments demonstrate that visual statistical learning is often altered depending on contextual factors that would be expected to fluctuate in everyday contexts.