Use this URL to cite or link to this record in EThOS:
Title: Exploring pattern separation as the mechanism underlying expectation effects on memory
Author: Frank, Darya
ISNI:       0000 0005 0289 8691
Awarding Body: University of Manchester
Current Institution: University of Manchester
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Access from Institution:
A discrepancy between expected and observed information often leads to better retention of the unexpected information. Despite extensive research, it is still unclear how unexpected information becomes more memorable. Therefore, the overarching goal of this thesis is to investigate how unexpected information is represented in memory. To achieve this, different sources of expectation were employed. First, expectation elicited by prior knowledge (schema) was examined using multi-element events, allowing manipulation of expectation orthogonal to schema incongruence. Unexpected incongruent items were more likely to be recalled than unrelated incongruent items, only in transitive inference and backward cued-recall, highlighting the sensitivity of expectation effects to hippocampal-dependent memory processes. Next, we examined whether a pattern separation (PS) computation, attributed to the hippocampus, underlies expectation-modulated memory. A stimulus database (SOLID) was created to allow tight control of perceptual similarity, which is key to tests of PS. The effect of contextual expectation (CE), a learned contingency between a cue and a semantic category, was examined using the similar foils. When CE was manipulated at encoding, better discrimination of unexpected highly similar foils was observed, with no difference between expectation conditions for targets and low similarity foils. A computational model simulating this task provided corroboration for the role of PS in CE effects, showing more dissimilar representations for unexpected information, driven by pattern-separated inputs from DG/CA3. Further evidence for the role of PS was found when CE was manipulated at retrieval. We observed a memory enhancement for unexpected targets and highly-similar foils, following their expected counterparts, coupled with increased activity in the hippocampus and midbrain dopaminergic regions. Finally, a novel source of expectation was identified, driven by previous mnemonic attribution, such that a current correct response was more likely following a previous incorrect response to a highly similar object from the same set. This expectation was associated with increased activity in PFC/ACC. Overall, the research reported here offers new insight into the mechanisms underlying different expectation effects on memory.
Supervisor: Montemurro, Marcelo ; Montaldi, Daniela Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: computational model ; hippocampus ; stimuli dataset ; pattern separation ; expectation