Learning from Memory Retrieval Attempts and Corrective Feedback

Abstract

Retrieving a memory can alter its form and its strength. Deliberate retrieval attempts take place often in everyday life, so understanding their effects is an important part of understanding the workings of memory as a whole. For example, deliberately retrieving information protects it somewhat against forgetting, and learning activities that involve retrieval (called retrieval practice) are often much more effective than similar activities that do not. The retrieval practice effect appears to be very general, having been found in a variety of tasks, educational subject areas, and age groups. In fact, learning benefits can be gained even when retrieval attempts fail—in many situations, incorrectly guessing an answer improves learning, as long as the correct answer is given to the learner after they guess. This thesis is intended to increase our knowledge of the cognitive mechanisms responsible for these effects, and of the practical value of retrieval practice in education.In Chapter 1, I explore whether corrective feedback effects, like standard retrieval practice effects, also depend on the strength of the relationship between the cue used to initiate a retrieval attempt (or guess) and the target information that is to be learned. Our findings indicate that relatively strong pre-existing cue-target relationships weaken both effects, giving some support to the idea that they arise from some common mechanisms.In Chapter 2, I report experiments using the corrective feedback paradigm to test the mediation hypothesis, a general mechanistic account that has been advanced to explain retrieval practice and corrective feedback effects. These experiments produced evidence against the mediation hypothesis when applied to learning from corrective feedback, and cast some doubt on its ability to explain retrieval practice effects in general. Further theoretical work may consider alternative explanations, perhaps involving roles for curiosity and surprise.In Chapter 3, I report an experiment with a retrieval practice and feedback activity in a school classroom. In running this experiment we were interested in how retrieval difficulty and student performance during practice might affect the reliability of the retrieval practice effect. Although our manipulations of retrieval difficulty had inconsistent effects, retrieval practice with feedback was a more effective review activity than reading for most students. This advantage held true even for students who answered most of their practice questions incorrectly, perhaps because of strong feedback effects, suggesting that properly implemented retrieval practice can help students regardless of their ability. Our procedures for ensuring prior exposure to the learning material and attention to feedback may have been important here, and these factors deserve careful consideration in future applied research.