AN INVESTIGATION OF THE IMAGINAL CODING AND RECALL LEVEL INTERPRETATIONS OF THE HYPERMNESIA PHENOMENON (IMAGERY, MEMORY)

DAVID GRAHAM PAYNE, Purdue University

Abstract

Hypermnesia (increased recall levels across an increasing retention interval) was investigated in seven experiments designed to test two theoretical accounts of the phenomenon. The imaginal coding hypothesis (Erdelyi & Becker, 1974) argues that imagery is critical in producing hypermnesia. The recall level hypothesis (Roediger, Payne, Gillespie & Lean, 1982) states that hypermnesia is related directly to the recall level across conditions in an experiment. These hypotheses were tested directly using picture and word lists and manipulating recall level by varying the item presentation rate (Experiments 1 & 2) or the number of list presentations (Experiment 3). Results showed that when recall levels were equated (Experiments 1 & 3), pictures still tended to produce a larger hypermnesic effect than words. Imagery, however, is not critical in producing hypermnesia, since hypermnesia was also obtained with words. (Also in a semantic memory task in Experiment 4.) Semantic (Experiment 4) and episodic (Experiment 5) memory tasks were used to test the recall level prediction that greater hypermnesia should be obtained with short tests than long tests. Results showed that hypermnesia was not related to test length. Experiment 6 tested the notion that hypermnesia is the result of subjects producing covert retrieval cues. Analyses of item recovery patterns obtained with a categorized word list generally supported this self-cuing hypothesis. Experiment 7 tested the hypothesis that the picture-word difference is due to differential learning during testing for these two item types. Attempts to vary recoding at the time of testing provided only weak support for this learning hypothesis. The final section of the paper presents alternative theoretical interpretations of the hypermnesia phenomenon.

Degree

Ph.D.

Subject Area

Psychology|Experiments

COinS