Mechanisms of Autocorrelation in Recognition Memory

Models of recognition memory typically assume that decisions are made independently from one judgment to the next. However, a growing body of evidence suggests that recognition decisions exhibit sequential dependencies, where information from the previous judgment carries forward in time to influence subsequent decisions. Across three chapters, this thesis examines these dependencies and their underlying mechanisms. In Chapter 1, we begin by conceptually replicating the established response-based carryover effect, characterized by the perseveration of memory judgments across subsequent trials. We then demonstrate a novel content dependency where both perceptual (orthographic) and conceptual (semantic) similarity between consecutive test items increase the likelihood of judging an item to be “old”. Lastly, Chapter 1 uses a manipulation of attentional engagement to challenge attention-based accounts of sequential dependencies. Chapter 2 disentangles decision repetition from motor priming by dissociating judgment and motor response. Our findings reveal that decision perseveration persists even when motor repetition is eliminated, supporting the notion that carryover effects are rooted in decision processes, rather than motor planning. Finally, in Chapter 3, we use computational modelling to test two potential mechanisms for sequential dependencies: Probe Transfer, through which consecutive item features are blended together before probing memory, and Familiarity Linger, through which the familiarity signal produced on a trial is incorporated into the subsequent judgment. The results of our simulations demonstrate that the blending together of item features across trials is able to capture both response and content dependencies. Thus, the work included in this thesis underscores that recognition decisions are not made independently. Additionally, we propose that a signal-change account of sequential dependencies that assumes item features persist across experiences may underlie response and content dependencies. These findings have implications for the methods used to measure memory and demonstrate the potential of using sequential dependencies to evaluate theories of memory.

Grantor dc:publisher

University of Guelph

Author dc:creator

• Dollois, Michelle, A.

Advisors dc:contributor.advisor

• Fiacconi, Chris M.
• Fenske, Mark J.

• recognition
• autocorrelation
• memory
• sequental dependencies

Statement dc:rights

• Attribution 4.0 International

Licence dc:rights.uri

• http://creativecommons.org/licenses/by/4.0/

Language dc:language.iso

en

Handle dc:identifier.uri

https://hdl.handle.net/10214/28808

OAI record oai:identifier

https://atrium.lib.uoguelph.ca/server/oai/request?verb=GetRecord&metadataPrefix=dim&identifier=oai%3Aatrium.lib.uoguelph.ca%3A10214%2F28808