Retrieval enhanced learning: Exploring the role of retrieval strategies

Phillip J Grimaldi, Purdue University

Abstract

Retrieval practice is an established and effective way to improve memory performance. A recent idea for why retrieval practice is so effective is that it helps an individual form a retrieval strategy, which aids in later recall. The purpose of this dissertation is to examine the relationship between retrieval strategies and retrieval practice. The dissertation examines whether retrieval strategies developed through retrieval practice contribute to retention, and whether different retrieval strategies moderate the mnemonic effects of retrieval practice. The experiments were designed to address these issues by using a part-list cuing procedure and by manipulating the retrieval strategies used during retrieval practice. In Experiment 1, subjects learned lists of words by practicing retrieval or by repeatedly studying. Subjects who practiced retrieval recalled more words after a delay than subjects who repeatedly studied. For both conditions, recall was negatively affected by part-list cues. However, when the part-list cues were removed, subjects who practiced retrieval showed a release from part-list cuing effect, whereas subjects who studied did not. According to the two-mechanism account of part-list cuing (Bäuml & Aslan, 2006), this finding suggests that retrieval practice resulted in the development of a retrieval strategy driven by interitem associations. Experiment 2 was designed to test whether disrupting the development of a retrieval strategy would reduce the mnemonic benefits of retrieval practice. Some subjects practiced retrieval in the same manner as Experiment 1, while others were forced to adopt a random output order during retrieval practice. Forcing a random output order successfully prevented the development of a consistent retrieval strategy, however this manipulation had no effect on retention, failing to support the idea that developing a retrieval strategy contributes to the mnemonic effectiveness of retrieval practice. Interestingly, both conditions displayed a release from part-list cuing effect, which was not predicted by the two-mechanism account of part-list cuing. Experiment 3 was designed to examine whether using different strategies during retrieval practice produces differential effects on memory performance. Some subjects were instructed to recall with a serial recall strategy, an alphabetic strategy, or were given standard free recall instructions. Serial recall and alphabetic recall resulted in faster learning than free recall. However, the advantage of serial recall over free recall was lost after a delay, whereas the advantage of alphabetic recall persisted. Serial recall is believed to rely heavily on interitem associations, thus the results contradict those of Experiment 1 and suggest that the buildup of interitem associations may not be beneficial for retention.

Degree

Ph.D.

Advisors

Karpicke, Purdue University.

Subject Area

Cognitive psychology

COinS