Advanced

Value-Based Memory Integration: The Asymmetry of Loss and Reward in Episodic Memory

Yngve, Adam LU (2019) PSYP01 20181
Department of Psychology
Abstract
Memory integration is how series of similar memories can merge into overlapping representations as one abstract model of behavior to anticipate present and future episodes – facilitating instrumental use of episodic experience for higher-order inferences. Previous research has showed that a novel memory of reward can travel backwards in time through episodic memory and generalize its contents to previous episodes by implicit association, although these were never directly linked to any valuable outcome. Using electroencephalography (EEG), the objective of this study was to track how decisions in the present moment can be explained by prior experience, specifically how people come to approach and avoid stimuli of asymmetrical values of... (More)
Memory integration is how series of similar memories can merge into overlapping representations as one abstract model of behavior to anticipate present and future episodes – facilitating instrumental use of episodic experience for higher-order inferences. Previous research has showed that a novel memory of reward can travel backwards in time through episodic memory and generalize its contents to previous episodes by implicit association, although these were never directly linked to any valuable outcome. Using electroencephalography (EEG), the objective of this study was to track how decisions in the present moment can be explained by prior experience, specifically how people come to approach and avoid stimuli of asymmetrical values of reward and loss as a function of memory integration over three experimental phases of i) association, ii) conditioning, and iii) decision. A central premise was that loss is more aversive than reward is attractive, in keeping with loss aversion. Behavioral results showed that participants (n = 26) generalized values in the predicted directions to previously familiar stimuli cues of reward and loss alike at decisions after direct value conditioning of their associates, albeit only at an individual level. Neural correlates of the outcome of directly rewarded stimuli over posterior scalp regions during conditioning were functionally associated with subsequent reward generalization at decision. The outcome of associates directly linked to punishments of loss elicited a more pronounced electrophysiological response than rewarded associates in overlapping posterior regions, yet these neural correlates did not account for generalization of loss. Implications and future directions are discussed. (Less)
Please use this url to cite or link to this publication:
author
Yngve, Adam LU
supervisor
organization
course
PSYP01 20181
year
type
H2 - Master's Degree (Two Years)
subject
keywords
memory integration, episodic memory, reward, loss, EEG, ERP
language
English
id
8967418
date added to LUP
2019-01-22 15:52:11
date last changed
2019-01-22 15:52:11
@misc{8967418,
  abstract     = {Memory integration is how series of similar memories can merge into overlapping representations as one abstract model of behavior to anticipate present and future episodes – facilitating instrumental use of episodic experience for higher-order inferences. Previous research has showed that a novel memory of reward can travel backwards in time through episodic memory and generalize its contents to previous episodes by implicit association, although these were never directly linked to any valuable outcome. Using electroencephalography (EEG), the objective of this study was to track how decisions in the present moment can be explained by prior experience, specifically how people come to approach and avoid stimuli of asymmetrical values of reward and loss as a function of memory integration over three experimental phases of i) association, ii) conditioning, and iii) decision. A central premise was that loss is more aversive than reward is attractive, in keeping with loss aversion. Behavioral results showed that participants (n = 26) generalized values in the predicted directions to previously familiar stimuli cues of reward and loss alike at decisions after direct value conditioning of their associates, albeit only at an individual level. Neural correlates of the outcome of directly rewarded stimuli over posterior scalp regions during conditioning were functionally associated with subsequent reward generalization at decision. The outcome of associates directly linked to punishments of loss elicited a more pronounced electrophysiological response than rewarded associates in overlapping posterior regions, yet these neural correlates did not account for generalization of loss. Implications and future directions are discussed.},
  author       = {Yngve, Adam},
  keyword      = {memory integration,episodic memory,reward,loss,EEG,ERP},
  language     = {eng},
  note         = {Student Paper},
  title        = {Value-Based Memory Integration: The Asymmetry of Loss and Reward in Episodic Memory},
  year         = {2019},
}