Neural effects of prioritisation on working memory

Thesis

Working Memory describes the cognitive ability to store information over short timescales, and use it to guide behaviour. It is strikingly limited, and we are only able to maintain small amounts of information in this short-term store. However, it is highly flexible and we are able to prioritise working-memory contents according to their relevance for behaviour. Such attentional prioritisation is associated with enhancements in working-memory performance. However, whether we use memories to guide behaviour should depend upon their quality and value. Thus, it is important to understand how we introspect upon working memory, evaluating and learning about its quality, and assign value to its contents. This thesis explores uncertainty and value in working memory. Chapter 2 examines how attentional prioritisation affects working-memory introspection. I show that attentional prioritisation improves the quality of introspection into working-memory contents. I also show evidence that feedback regarding the accuracy of recent confidence judgements is incorporated into working-memory confidence, shaping our beliefs about the quality of working-memory performance. Chapter 3 describes two experiments using electroencephalography (EEG) to explore how feedback regarding working-memory performance is processed. I show that uncertainty – manipulated externally by varying working-memory load, and carefully measured through subjective confidence judgements – is represented in the neural response to feedback. When feedback contains information regarding the quality of subjective uncertainty, I find evidence of a confidence prediction error that could facilitate learning about working-memory quality. Chapter 4 describes two novel experiments that explore how the value associated with objects held in working memory shapes performance. These experiments present reward information during working-memory maintenance to examine how reward influences the prioritisation of working-memory contents after encoding. I show that reward can guide the prioritisation of working-memory contents – even when reward is irrelevant to the task at hand. I demonstrate both global and item-specific benefits associated with reward-guided attentional prioritisation of working-memory contents. Finally, in Chapter 5, I discuss the findings of this thesis in a wider context, and remaining theoretical questions related to working-memory guided decisions.

Authors: Chekroud, SR

• DOI: 10.5287/ora-2z8n9jxwz
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: EEG; Reward; Attention; Memory; Feedback
• Subjects: Cognitive neuroscience; Psychology, Experimental; Psychology