How do students improve their value-based learning with task experience?

Abstract

When learning items that vary in reward, students improve their scores (i.e., earned reward) with task experience. In four experiments, we examined whether such improvements arise from better selective encoding of items that would earn more (vs. less) reward. Participants studied and recalled words across multiple study-test trials. On each trial, 12 words were slated with different values (typically from 1 to 12), and participants earned the point value assigned to a given word if it was correctly recalled. In all experiments, participants earned more points across the first two trials. In Experiment 1, participants either self-paced their study or had experimenter-paced study and in Experiment 2, some participants were penalised for each second spent during study. Improvements in points earned were related to increases in overall recall but not to selective encoding. In Experiment 3, some participants were given value-emphasised instructions, yet they did not demonstrate selective encoding. In Experiment 4, we used a larger range of point values, but selective encoding still did not account for the improvement in point scores across lists. These results suggest that metacognitively-driven selective encoding is not necessary to observe improvements in value-based learning.

Keywords:

• Value-based learning
• Selective encoding
• Learning how to learn
• Metamemory

This research was conducted as a senior thesis by the first author at Kent State University and was supported by the James S. McDonnell Foundation 21st Century Science Initiative in Bridging Brain, Mind, and Behavior Collaborative Award, and by the National Science Foundation Graduate Research Fellowship Program [grant number DGE-1321845], Amendment II. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

This research was conducted as a senior thesis by the first author at Kent State University and was supported by the James S. McDonnell Foundation 21st Century Science Initiative in Bridging Brain, Mind, and Behavior Collaborative Award, and by the National Science Foundation Graduate Research Fellowship Program [grant number DGE-1321845], Amendment II. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

Notes

¹ Because Experiment 4 used values from discrete ranges of values (higher and lower), some estimates of the selectivity index would fall outside the interpretable range of −1 to +1 (Watkins & Bloom, 1999).