Researcher ORCID Identifier
0009-0003-9499-9301
Graduation Year
2026
Date of Submission
12-2025
Document Type
Open Access Senior Thesis
Degree Name
Bachelor of Arts
Department
W.M. Keck Science Department
Reader 1
Zeynep Enkavi
Reader 2
Xiao Zhang
Terms of Use & License Information
Rights Information
@ 2025 Allison E Rudolph
Abstract
Learning from experience requires integrating feedback over time, yet individuals vary widely in how effectively they extract structure from probabilistic environments. This study examined how such variability relates to neural responses in the ventral striatum (vStr), caudate nucleus, hippocampus, and medial prefrontal cortex (mPFC). Seventy-four participants across childhood, adolescence, and adulthood completed a modified Iowa Gambling Task during fMRI scanning, choosing whether to play or pass on four machines differing in expected value and variance. Behavioral learning slopes, calculated across six runs for each machine, were used to classify participants as learners or non-learners. Learners showed clearer improvement over time and earned more points overall, whereas non-learners responded more slowly and performed especially poorly on negative expected-value machines.
Neuroimaging analyses focused on outcome valence, reward magnitude, and event-type contrasts. Learners exhibited stronger vStr responses to positive versus negative feedback and to win–loss contrasts, indicating enhanced sensitivity to rewarding outcomes. In contrast, the caudate showed no learner-related differences in valence or magnitude contrasts and demonstrated greater win–loss responsivity in non-learners, suggesting heightened reactivity to immediate outcomes without integration into a stable choice policy. Outside the striatum, both the hippocampus and mPFC showed robust event-type effects, with learners exhibiting greater differentiation between stimulus and feedback processing. This pattern implies more structured encoding and more effective use of feedback information.
Together, these results identify a functional dissociation across frontostriatal and hippocampal–prefrontal systems. Successful learning was associated with enhanced reward signaling in the vStr and greater event-specific differentiation in hippocampus and mPFC, whereas non-learners showed more reactive but less integrated neural responses.
Recommended Citation
Rudolph, Allison E., "Neuroimaging Correlates of Differences in Learning from Repeated Probabilistic Outcomes" (2026). CMC Senior Theses. 4251.
https://scholarship.claremont.edu/cmc_theses/4251
