Date of Award

Spring 1-1-2014

Document Type


Degree Name

Doctor of Philosophy (PhD)


Psychology & Neuroscience

First Advisor

Alice F. Healy

Second Advisor

Tim Curran

Third Advisor

Matt Jones

Fourth Advisor

Mike Mozer

Fifth Advisor

Jerry Rudy


Feedback informs learners whether or not a representation should be changed, and directs learners’ attention towards information that can be used to modify or refine that representation. What sorts of information are better at accomplishing these objectives in explicit learning than others? Under what circumstances? In order to help answer these questions and contribute to the understanding of how feedback information guides learning, the experiments in this thesis investigated both the individual and combinatory effects of different variants of performance (e.g., performance correctness) and task (e.g., providing the correct answer) directed feedback information. Given that many feedback environments are group-based, we compared performance feedback based on both an individual’s and a group’s performance in order to understand how such information differentially impacts individuals’ learning. The first two experiments utilized a fact-learning task, with the focus of the first experiment on both the individual and combinatory effects of correct answer, individual performance, and group performance feedback, and the focus of the second experiment more specifically on group performance feedback. The last experiment utilized a more complex rule-based explicit learning task, and focused on the isolated effects of group performance, individual performance, correct answer, and a different sort of task information, rule feedback. The results of all experiments suggest that feedback containing task information is more effective than feedback containing performance information, but that more explanatory rule feedback may be more cumbersome to process than correct answer feedback (Experiment 3). When feedback contains multiple pieces of information, the efficacy of feedback is reduced when the contents are not complementary in where they cue attention, specifically when one piece cues attention towards but the other away from task information (Experiments 1 & 2). Further, the results suggest that even though performance information does not directly state task information, it can cue attention towards productive task-processing, but that the benefits derived from such cueing are primarily evident when task information is immediately available for processing (Experiment 3) and, if the performance information is paired with task information, when those two contents are congruous in this mission (Experiment 1 & 2).