Interactions between reward and visual statistical learning
Date
2020
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Humans are adept at learning regularities in a visual environment, even without explicit cues to structure and in the absence of instruction—this has been termed “visual statistical learning” (VSL). Previous studies provide evidence that some aspects of learning from reward resemble statistical learning, yet whether and how reward learning impacts VSL is largely unexamined. In this dissertation, we examine the impact of reward information and learning on VSL with two different questions: 1) how is VSL modulated by reward when reward information is embedded within visual regularities (Chapter 2 and 3), and 2) how is the representation generated by VSL impacted by novel reward information (Chapter 4)? First, we found that reward contingencies affect VSL, with high-value associated with stronger behavioral and neural signatures of such learning than low-value images. Specifically, differences between high-value first images and low-value first images were consistent with differences in attentional engagement, suggesting that the powerful allocation of attention in response to the high-value first image potentially enables better memory for statistically learned pairs and reward information than the low-value first image. Secondly, we investigated the nature of the representation resulting from standard VSL and how it is shaped by reward. Our work supports the idea of generalized and flexible representation in VSL and provides evidence about how this representation is structured. Next, when learned statistical structured information was cued by novel reward information, this post reward cue retroactively refreshed the previous memory and enhanced incidental learning. Additionally, a reward cue led to the higher flexibility of the representation resulting from VSL than no reward cue. In sum, this dissertation demonstrates novel evidence that VSL is modulated by reward information, and highlights evidence for a possible mechanism. Reward information may shape VSL due to its influence on the allocation of attention to stimuli during learning and via its role of shaping memory consolidation retroactively.
Description
Keywords
Reward learning, Visual statistical learning