Dissertations, Theses, and Capstone Projects

Date of Degree

6-2026

Document Type

Master's Thesis

Degree Name

Master of Science

Program

Cognitive Neuroscience

Advisor

Jennifer Mangels

Subject Categories

Behavioral Neurobiology | Cognitive Neuroscience | Cognitive Psychology | Oncology | Psychology

Keywords

neuroscience, electroencephalogram, event-related potential, breast cancer, metamemory, recognition memory

Abstract

Breast cancer survivors (BCS) often report noticeable cognitive difficulties despite showing limited impairment on standardized neuropsychological tests. Some have theorized that this is simply an insensitivity of standardized testing to subtle cognitive changes (test-insensitivity hypothesis), whereas others have suggested there are early neural changes that are detected by patients but precede detection by these tests (early signal hypothesis) or that when they achieve normal performance, it is through saliently effortful compensatory processes (compensation hypothesis). The present study tested these theories by assessing metamemory and recognition memory retrieval in BCS and non-cancer controls (NCC) using both behavioral and event-related potential (ERP) measures, the latter of which might reveal BCS-related changes in neural mechanisms even when behavior appeared similar to NCC. During electroencephalography recording, participants completed a visual study-test word recognition task in which items were encoded either under distraction (background Japanese audio), thus providing an encoding challenge, or non-distraction (no audio) conditions. During retrieval, participants made old/new recognition judgments and retrospective confidence judgments (RCJs). Behavioral analyses assessed discrimination sensitivity, response bias, and retrospective confidence indices.  Event-related potential (ERP) analyses examined retrieval-related activity at frontal and parietal sites across mid-latency (300-500 ms) and later (500-700 ms) time windows. Although distraction reduced both item and confidence discrimination in both groups, BCS and NCC showed comparable discrimination performance regardless of condition. However, relative to NCC, BCS showed a less liberal response bias and greater RCJ underestimation. ERP analysis confirmed a frontal old/new effect from 300-500 ms, but no reliable group differences. However, ERPs during 500-700 ms window demonstrated greater frontal positivity and parietal negativity in BCS compared to NCC, regardless of condition or item type (old, new). However, there were additional old/new effects that varied by group and condition. While these results confirmed a discrepancy between objective (discrimination) and subjective (confidence) measures, the presence of different patterns of later ERP activity (500-700 ms), combined with largely intact memory retrieval even under more challenging encoding conditions (i.e., distraction) and typical mid-latency (300-500 ms) old/new effects,  suggests differences in post-retrieval processing that are most aligned with the compensatory explanation. Overall, these findings indicate a discrepancy across subjective, behavioral, and neural measures and support the view that standard behavioral measures may not fully capture meaningful cognitive differences in breast cancer survivorship.

Download

Share

COinS