Dissertations, Theses, and Capstone Projects

Date of Degree

6-2026

Document Type

Master's Thesis

Degree Name

Master of Science

Program

Cognitive Neuroscience

Advisor

David Johnson

Committee Members

Tony Ro

Subject Categories

Behavior and Behavior Mechanisms | Cognitive Science

Keywords

Fear, Fear Learning, Uncertainty

Abstract

The ability to discriminate between cues predictive of threat vs. safety (fear acquisition) and subsequently update fear responding when contingencies change (fear extinction) is crucial to survival for all organisms. While these processes have been the subject of intense experimental scrutiny for several decades, most human fear learning studies have focused primarily on extinction learning and fear recovery processes, with less focus on fear acquisition itself. Indeed, researchers typically aim to minimize variability during fear acquisition in order to rule out differences in extinction being driven by variability during the initial learning phase. Thus, experimental protocols are often optimized to maximize differences in response magnitude between threat vs. safety cues. While this is a reasonable approach, this emphasis may have inadvertently reduced sensitivity to capture individual and group differences in fear acquisition, shaping the view that acquisition yields fewer informative effects than other phases of fear learning.

Here, we used a protocol optimized to increase variability in acquisition and potentially increase sensitivity to capture individual and group differences during this phase. We aimed to test two factors – biological sex and three trait measures of negative emotionality, specifically the  State-Trait Anxiety Inventory, Intolerance of Uncertainty, and the Big Five Inventory – as potential mediators of threat-safety discrimination learning. Participants (n = 21) were presented with three cues – a non-reinforced safety cue (CSM), a threat cue associated with low uncertainty (CSP1) and a threat cue associated with high uncertainty (CSP2). Participants completed three fear-relevant trait measures via self-report, Intolerance of Uncertainty, the Big Five survey, and the State Trait Anxiety Inventory. Female participants reported the number of days since the onset of their menstrual cycle, allowing classification into low vs high estrogen groups, according to typical cycle-related fluctuations in estrogen. Our primary measure of learning was skin conductance response (SCR). Results show a significant effect of time on learning, F (1, 20) = 7.058, p = .015, η² = 0.037, but not stimulus, F (2, 40) = 0.220, p = 0.803, η² = 0.003. There were no significant effects for the trait measures. While there was no significant association between sex/sex hormone levels and discrimination learning, as indexed by SCR, low-estrogen females reported higher state anxiety both before and after acquisition compared to high-estrogen females and males, F (2, 16) = 8.50, p = 0.003, n2 = 0.468, consistent with previous reports. It should be noted that the present design—incorporating three interleaved cues with varying levels of uncertainty—is more complex than those typically used in human fear-learning studies, introducing greater variability and thus requiring a larger sample size than simpler paradigms. Overall, these findings underscore both the promise and challenges of studying fear acquisition using paradigms characterized by high complexity and uncertainty. Because the present study is underpowered at this stage, continued data collection will be critical for evaluating the emergence of additional effects.

Download

Share

COinS