Abstract

This dissertation investigates the architecture of reward-based learning in conflict tasks, examining the robustness of item-level reward associations and whether reward learning extends beyond item features to spatial contexts, visual properties, and relational features. Across five pre-registered experiments, participants discovered reward associations through trial-by-trial feedback, and we examined whether these associations persisted when rewards were subsequently removed. Experiment 1 demonstrated that item (color)-reward associations learned during a Stroop task transferred robustly to a flanker task in the absence of reward, suggesting that item-reward associations form relatively abstract, task-independent value representations that influence attention across contexts. Experiments 2 through 4 probed whether reward learning extends to non-item features, including spatial context (Experiment 2), stimulus shape (Experiment 3), and congruency (Experiment 4). Across all three experiments, participants failed to acquire reward associations for these features despite their predictive validity. Experiment 5 placed item and non-item (location) features in direct competition, revealing that participants selectively acquired item-reward associations even when alternative features were equally informative. These results point to a hierarchical organization of value-based learning, with item identity holding privileged status over non-item features in forming reward associations. The privileged status of item identity in reward learning may reflect the direct link between item features, task-relevant responding, and reward outcomes. These findings establish item identity as a fundamental organizing principle of value-based attention, revealing that reward learning is selectively tuned to task-relevant item features.

Committee Chair

Julie Bugg

Committee Members

Jonathan Bogard; Richard Abrams; Todd Braver; Wouter Kool

Degree

Doctor of Philosophy (PhD)

Author's Department

Psychology

Author's School

Graduate School of Arts and Sciences

Document Type

Dissertation

Date of Award

4-24-2026

Language

English (en)

Available for download on Thursday, April 22, 2027

Share

COinS