Imagery and organization in memory: Instructional effects

Abstract

This paper concerns the relation between mental organization and memory performance. In a series of experiments, it is determined that interactive imagery exceeds separate imagery in cued recall and the organization of free recall. However, the two are indistinguishable in either recognition or the level of free recall unless some additional interunit organization is initially encoded. Although interactive imagery allows a list to be retained as fewer traces than does separate imagery, those traces are necessarily larger. The inverse relation between the number and size of traces is found also in the organization of free recall. This tradeoff results in equivalent total recall.There is no shortage of evidence that imagery instructions are correlated with memory performance. However, the contribution of the research to the understanding of the memory system has been disappointing. For example, Runquist (1973) summarized the impact of the research as "look what we can do with imagery instructions" (p.295). In a similar vein, Tulving and Madigan (1970) questioned the value of using imagery as an explanatory construct and called for more integration of imagery research with memory theory in general. By 1975, Postman detected some progress along these lines, but concluded that the theoretical efforts of Begg (1972Begg ( , 1973 and Paivio (1971Paivio ( , 1975a were inching forward slowly at best. The present paper offers a memory-theoretic account of the effects of imagery instructions, with the hope of adequately describing those effects in terms that relate the research to general memory processes. In brief, imaging items together, as opposed to separately, reduces the number of memory units required to retain a given list by increasing the average size of the units. However, any organizational format will be most appropriate for some contexts of use, but will be irrelevant or even antagonistic for other purposes. Consequently, the present paper will be as much concerned with what imagery instructions do not accomplish as with what they do.Before casting imagery instructions in theoretical terms, let us review the effects of the instructions, based on previous research. In the basic situation of interest, subjects are presented with a list of familiar concrete nouns, randomly paired with each other.The research reported here was supported by Grant A8122 from the National Research Council of Canada. I thank the members of the F. T. Bacon Society (Grant Harris, John Mitterer, and Doug Upfold) for considerable support. Reprints can be obtained from Ian Begg, Department of Psychology, McMaster University, Hamilton, Ontario, Canada L8S 4Kl.One group is instructed to form a single image to represent each pair of items interacting in some way, while a second group is instructed to form a separate image for each member of each pair. In subsequent tests of memory, some results are quite intuitive and are easily accommodated by any theory of memory. For example, if one member of each pair serves as a cue for recalling the ot...

Role of contextual imagery in associative recall

“…Subjects who employed separation imagery showed no increase in associative recall compared to a verbal repetition control, while an interactive image group showed a much higher level of performance. Begg (1973Begg ( , 1978 and Robbins, Bray, Irvin, and Wise (1974) found a similar outcome with concrete words, as did Neisser and Kerr (1973) when the words were embedded in sentences describing either interactive or separate pictorial frames. The research reported here shows that the mnemonic effects of separation imagery are intertwined with the imagined context in which the images of the word referents are set.…”

“…Bower (1970Bower ( , 1972 interprets the superior performance he obtained with interactive imagery along much the same lines, saying, "I am convinced that associating meaningful terms is largely a matter of relating them in some organization" (1972, p. 86), a thesis that, he points out, has been advocated by Gestalt thinkers such as Asch (1969). Begg (1973Begg ( , 1978 has offered a sirnilar interpretation.…”

“…This confounding with spacing is equal for both interactive and separation imagery. Subjects were not explicitly instructed to keep successive images separate from each other in the manner described by Begg (1978). An unanticipated cued recall test followed immediately after presentation of the last pair.…”

Similarity and contrast in memory for relations

“…By the present account, successful performance requires that the cue be encoded, that the encoded cue contact or access its parent trace, and that the trace contain or provide sufficient data to allow response production (cf. Begg, 1978;Martin, 1967). Recall could fail because a cue contacts the wrong trace or none at all, or because a contacted trace leads to production of the wrong response or none at all.…”

Beyond category sorting and pleasantness rating: Inducing relational and item-specific processing

“…We chose the six procedures from prior research that claimed or seemed to claim that the procedures induced relational or item-specific processing. The three relational procedures and their source were category sorting (Einstein & Hunt, 1980;Hunt & Einstein, 1981), narrative construction (Bower & Clark, 1969), and relational imagery (Begg, 1978;Bower, 1970;McGee, 1980;Schwartz & Humphreys, 1974). The item-specific procedures and their source were pleasantness rating (Einstein & Hunt, 1980;Hunt & Einstein, 1981;Packman & Battig, 1978;Toglia & Battig, 1978), familiarity rating (Packman & Battig, 1978;Toglia & Battig, 1978), and single imagery (Begg, 1978;Bower, 1970;McGee, 1980).…”