Do the electrophysiological correlates of recognition memory change with time?

  • Jenna Roberts

    Student thesis: Phd


    AbstractThe aim of this PhD thesis was to explore the way in which recognition memory changes over time. Of particular interest was how forgetting and systems level consolidation processes alter the qualitative nature of recognition judgments and the impact this has on event related potentials (ERPs) recorded during recognition. An emerging body of fMRI and animal work has started to suggest that changes to the neural basis of memory can be observed after intervals in the order of days and weeks. Although much research has examined the ERP correlates of recent recognition, there has been little attempt to compare this to remote recognition. This gap in the literature is investigated in the present PhD thesis over five ERP and two behavioural experiments. The first set of experiments investigated recognition success i.e. the subjective awareness that a stimulus has been encountered before. Previous work has associated familiarity-based recognition with an early midfrontal ERP effect whereas recollection-based recognition has been linked to a later onsetting parietal ERP effect. These effects were compared for recently studied stimuli and stimuli studied 1 week earlier. Results revealed an attenuation of the late parietal effect. This quantitative difference suggests that the neural networks underlying the ERPs for recent and remote recollection remain the same after a 1 week delay but may be less active after a period of forgetting. Behavioural work linked this to a drop in strength and episodic detail for remote recollection. Examining the midfrontal effect over time produced a more complex pattern of results. The effect was not reliable in Experiment 1 for remote familiarity judgments. In follow up experiments, however, midfrontal effects were reliable for week old memories but were not modulated by either delay or memory strength manipulations. These findings do not permit strong conclusions regarding the way familiarity memory and midfrontal ERPs vary over time, other than to say that the midfrontal effect is not a short lived correlate of recognition memory. A second set of experiments investigated how the length of the delay between study and test impacts on retrieval orientation and retrieval effort. Retrieval orientation refers to the way in which participants strategically alter how a memory cue is processed based on current task demands. Results showed ERP differences when remote memories were cued compared to when recent memories were cued. However, these differences were eliminated when recent and remote memory was matched for performance. This finding indicated that effort required at retrieval rather than memory age per se influenced differential cue processing. A follow up experiment supported this claim but found that participants may utilise delay information during recognition in a very general sense. In this experiment, ERPs indicated different cue processing when participants knew the age of the memories they were trying to retrieve compared to when they did not. Taken together, the results suggest that memory age does not influence the ERP correlates of recognition in a direct manner. More important are the indirect changes occurring over time, such as reduced memory strength, reduced episodic detail, slower reaction times and increased effort at retrieval. These variables, and possibly a range of others, should be controlled as far as possible in studies aiming to explore neural changes driven by systems consolidation. Prospective studies, where the learning phase can be controlled makes this aim feasible, as some of the experiments in this thesis have shown. Future work might benefit from focusing more on variations at encoding, rather than large differences in the length of the retention interval, as has been the traditional approach to this issue.
    Date of Award1 Aug 2013
    Original languageEnglish
    Awarding Institution
    • The University of Manchester
    SupervisorAndrew Mayes (Supervisor) & Dimitrios Tsivilis (Supervisor)


    • Recognition memory
    • Consolidation
    • Event related potential (ERP)

    Cite this