Elaboration In Memory Research Paper

2. Historical Antecedents Of The Construct Of Elaboration

The modern concept of elaboration emerged from a series of information-processing models emphasizing the roles of filtering and rehearsal of information as critical determinants of memory performance (Broadbent 1958, Atkinson and Shiffrin 1968, Baddeley and Hitch 1974). Filter models of attention (Broadbent 1958) emphasize that attention plays an important role in sensory processing by giving priority to information arriving through certain sensory modalities or channels while filtering information arriving through other modalities or channels. These models introduced the ideas that earlier stages of information processing were required for later processing and that only a limited subset of information received this initial processing. These notions of filtering were adopted in stage models of memory (Atkinson and Shiffrin 1968). Stage models assume three distinct types of memory stores (a sensory store, a short-term memory store, and a long-term memory store) and that information flows sequentially from one store to the next. Information entering the sensory store was assumed to be filtered by attentional processes so that only a limited set of this information entered the short-term memory store. Upon entrance to the short-term store, an intentional rehearsal process, generally referred to as maintenance rehearsal, was assumed to maintain information, mediating its transfer to the long-term store. Transfer to the long-term memory store was assumed to be a necessary condition for later memory performance. Thus, stage models emphasize that conscious, intentional processing is a critical influence on memory (see Baddeley and Hitch (1974) for more complex descriptions of intentional processes that might operate in short-term or working memory).

The emphasis of stage models on maintenance rehearsal led to a series of investigations of the properties of maintenance rehearsal. In the current context, the most critical studies were those which demonstrated that, contrary to the assumptions of the stage model, maintenance rehearsal did not necessarily enhance memory (Craik and Watkins 1973, Woodward et al. 1973). For example, Craik and Watkins (1973) demonstrated that increasing the duration of maintenance rehearsal by manipulating characteristics of a secondary task did not enhance memory. This counterintuitive finding led to a distinction between type 1 (maintenance) and type 2 (elaborative) rehearsal, with only the latter type of rehearsal assumed to affect retrieval from long-term memory. Thus, by the early 1970s, filter models of attention and stage models of memory had emphasized the role of conscious, intentional rehearsal processes in memory and ensuing empirical investigations had demonstrated that maintenance rehearsal did not necessarily enhance memory.

3. Specific Genesis Of The Modern Construct Of Elaboration In The Levels-Of-Processing Literature

Filter models of attention and stage models of memory focused attention on the process of memory storage. In the early 1970s, the levels-of-processing approach presented an alternative characterization of memory storage (Craik and Lockhart 1972). In contrast to stage models, the levels-of-processing approach assumed that to-be-remembered information was processed in three sequential stages of perceptual analysis.

Presented materials were assumed to be subjected to an orthographic analysis, a phonological analysis, and then a semantic analysis. As in stage models, the character and nature of this processing were assumed to influence memory. Specifically, it was assumed that earlier stages of analysis (i.e., orthographic and perceptual analyses) led to transient memory traces, whereas semantic analyses led to longer-lasting memory traces. Thus, the levels-of-processing framework predicts that semantic processing should lead to better memory than phonological or orthographic processing. Consistent with this prediction, numerous empirical studies demonstrated that semantic orienting questions (e.g., ‘is a pear a fruit?’) during study lead to better memory than phonological or orthographic orienting questions (e.g., ‘does pear rhyme with fair?’) during study. Despite this support for the levels-of-processing approach, the congruity effect (Craik and Tulving 1975) presented a major challenge to this approach. The congruity effect is the finding that memory is better when responses to orienting questions are positive (i.e., ‘yes’) rather than negative (i.e., ‘no’), independent of the level of processing of the orienting question. For example, the word ‘pear’ is better remembered if the orienting question is ‘Is a pear a fruit?’ than if the orienting question is ‘Is a pear a weapon?’ The congruity effect presents a problem for the levels-of-processing approach because it demonstrates substantial variability in memory performance with a fixed level of processing.

To explain the congruity effect, Craik and Tulving (1975) argued that elaboration—the association of to-be-remembered information with other information—was critical to memory performance. Thus, the congruity effect was assumed to occur because ‘positive’ orienting questions created episodic associations between the to-be-remembered item and information in the orienting questions (e.g., ‘pear’ and ‘fruit’) whereas ‘negative’ orienting questions did not form such associations (e.g., ‘pear’ and ‘weapon’). Further, Craik and Tulving (1975) proposed that elaboration was a fundamental determinant of memory performance, hypothesizing that it was the basis of levels-of-processing effects. Specifically, they argued that the elaborations produced by semantic orienting questions were more likely to enhance memory than the elaborations produced by phonological or orthographic orienting questions.

4. Theoretical Mechanisms Of The Effect Of Elaboration On Memory Retrieval

Two classical approaches to memory retrieval provide explanations of elaboration’s effects on memory. These mechanisms are compatible and one can conceptualize effects of elaboration as arising from their joint influence. The first mechanism arises from associationist models of memory (e.g., Raaijmakers and Shiffrin 1981). In these models, to-be-remembered information is associated with episodic or list representations and to other information presented or retrieved during a study episode. During a memory test, episodic and other representations can be retrieved and used as cues for to-be-remembered information. In this perspective, elaboration enhances memory because it creates additional cues for to-be-remembered information. A simple example demonstrates the benefit of additional cues. The presentation of the word ‘dog’ in the study list creates an association between this item and the representation of the study list. If a subject then elaborates the presentation of the word ‘dog’ by recognizing that ‘dog’ is associated with ‘cat,’ this elaboration creates a second cue (i.e., ‘cat’) for ‘dog.’ In addition to providing an account of the effects of elaboration, a focus on multiple cues also provides an account of other memory effects such as those of encoding variability and imagery. Encoding variability effects occur when information studied from two perspectives is remembered better than information studied twice from a single perspective (Reder et al. 1974). For example, studying the word ‘bank’ in relation to its meanings as ‘the side of a river’ and ‘a financial institution’ produces better memory than studying ‘bank’ twice in the context of either of the individual meanings. Effects of imagery occur when words that have concrete referents (e.g., ‘dog’) are remembered better than words with abstract referents (e.g., ‘justice’) (Paivio 1971). The effects of encoding variability and imagery can be ascribed to the availability of additional retrieval cues. In the case of encoding variability, the second meaning of a word provides an additional cue, while in the case of concrete words, a visual imaginal representation provides an additional cue.

The second mechanism focuses on the idea that elaboration creates distinctive associations (Craik and Jacoby 1979). This approach assumes that cues are more effective when they have unique associations to to-be-remembered information. Thus, elaboration can enhance retrieval by creating distinctive or unique cues. For example, if ‘xylophone’ is part of a to-be-remembered list, the elaboration ‘a musical instrument beginning with x’ should enhance performance more than the elaboration ‘a word containing many vowels.’

This focus on distinctiveness has been used to explain the standard levels-of-processing effect (i.e., semantic>phonological>orthographic) by assuming that semantic cues are more distinctive than phonological or orthographic cues. Further, investigators working in this tradition have demonstrated that nonsemantic information can be well remembered if it is distinctive. For example, Hunt and Elliott (1980) have demonstrated that orthographic distinctiveness can enhance recall and recognition.

The preceding account of the effects of elaboration emphasizes the cue-dependent nature of memory. The availability and distinctiveness of cues are critical determinants of memory performance. This focus on cue dependence helps explain how powerful mnemonic devices such as hierarchical organization (Bower 1970) produce exceptional memory performance. In hierarchical organization, each cue is associated with a relatively small number of to-be-remembered items which, in turn, serve as cues for other items. Thus, the initial retrieval of a single cue at the top of a hierarchical organization can permit access to a large amount of information. Consistent with this view, hierarchical organization has been implicated in the exceptional performance of many mnemonists (e.g., Chase and Simon 1973).

5. Applications Of The Elaboration Construct

5.1 Interactions With Populations

Difficulties in using elaborations have been proposed as explanations of differences in memory performance across populations. Most notably, difficulties in using elaborations have been proposed to explain why both young children and older adults have poor memories. Consistent with this account, studies of young children have demonstrated that young children are much less likely to create spontaneously elaborations than older children. For example, Barnes et al. (1996) demonstrated that six-year olds were less likely to construct elaborations about a text than 15-year olds, even when the two groups of subjects were equated on prior knowledge of the text. Similar arguments have been applied to explain the memory deficits of older adults. In addition, it has been proposed that the elaborations produced by older adults are more idiosyncratic than those produced by younger adults. These idiosyncratic elaborations are unlikely to be effective cues when they are retrieved during a later test. Consistent with this view, Zabrucky and Moore (1995) demonstrated that elaborations of text generated by older adults were less likely to be related to the text (i.e., more idiosyncratic) than the elaborations generated by younger adults. Further, the number of idiosyncratic elaborations generated was negatively correlated with recall and accounted for much of the variance in recall performance across older and younger subjects.

Additional evidence that deficits in elaboration produce the memory problems demonstrated by young children and older adults is found in studies manipulating elaborative processing. These studies demonstrate that manipulations designed to increase elaboration minimize the memory deficits demonstrated by younger children and older adults. For example, Cherry et al. (1993) demonstrated that asking older adults to create elaborations that focused on explaining a text minimized age differences in recall, even though such differences occurred in a control condition in which participants were not directed to construct elaborations.

The elaboration construct has also been applied to the explanation of memory difficulties in a wide variety of patient populations (i.e., learning disabled, brain injured, and mentally retarded persons). As with younger children and older adults, it is assumed that these patients do not spontaneously or effectively use elaborations. In this context, numerous studies have demonstrated that teaching elaboration techniques to these patients enhances their learning and memory. For example, Scruggs et al. (1994) demonstrated that fourth and fifth graders with mild learning disabilities demonstrated a substantial benefit in text recall when encouraged to construct elaborations.

5.2 Interactions With Materials And Prior Knowledge

A large number of studies have investigated the interactions of elaborative processing with the nature of the to-be-learned materials and the participant’s prior knowledge. These investigations help specify the conditions under which elaborations enhance memory. The semantic relatedness of to-be-learned materials is one of the most important determinants of memory with strongly related materials generally being remembered better than weakly related materials. Further, the effects of elaboration interact with those of semantic relatedness, so that elaborative processing produces a larger benefit for weakly related than strongly related materials, often equating overall memory performance on these materials. For example, Horiba (1996) examined recall of pairs of sentences that were either strongly or weakly related when participants were given elaboration or memorization instructions. Horiba found that strongly related sentences were remembered better than weakly related sentences under memorization instructions, but that elaboration produced a greater benefit for the weakly related sentences, equating performance on strongly and weakly related sentences. Explanations of such finding emphasize that elaborations occur spontaneously for strongly related materials, but not for weakly related materials. Consequently, elaborative instructions produce a greater benefit on weakly related materials.

This focus on the beneficial effects of elaborative processing for weakly related materials can be used to explain circumstances in which weakly related materials produce better memory than strongly related materials. For example, Erdfelder and Bredenkamp (1998) hypothesized that atypical events presented in text are better remembered than typical events because subjects must extensively elaborate the atypical events to produce a coherent mental representation of the text.

One assumption underlying the preceding account of the beneficial effects of elaboration on weakly related materials is that participants can retrieve the semantic information necessary to create the appropriate elaborations. This assumption suggests that participants’ degree of prior knowledge should interact with the effects of elaborative processing. Consistent with this suggestion, numerous studies have demonstrated that elaborations produce greater memory benefits when participants have more extensive prior knowledge. Willoughby et al. (1993) demonstrated this finding by examining memory for facts learned about well-known and unfamiliar animals. They compared memory when subjects were asked to explain why a fact about an animal was true to memory in a repetition control condition. The elaboration condition produced greater memory relative to the control condition for familiar animals, but not for unfamiliar animals. The absence of an elaboration effect for unfamiliar animals is consistent with the view that prior knowledge is critical to the beneficial effects of elaborative processing.

These interactions of elaborative processing with the structure of materials and participants’ prior knowledge play an important role in memory for, and comprehension of, text (Kintsch and Van Dijk 1978). While an extensive discussion of the text comprehension literature is beyond the scope of this research paper, there are two important parallels between the literature on text comprehension and memory that should be noted. First, consistent with the views derived from standard memory paradigms, elaborations of a text are assumed to have important effects on the memory representation of the text. Second, the issue of the materials and knowledge states that permit elaborations has been a topic of extensive study. One example that represents some of the issues arising in such investigations is Lea’s research (Lea 1995) on inference generation. Lea demonstrated that subjects will use propositional logic (e.g., given p or q and not p, they will infer q) to generate elaborations about a text, even when such elaborations are not necessary to maintain a coherent text representation.

5.3 Interactions With Memory Tests

As discussed above, numerous studies have demonstrated that elaborating materials during study enhances performance in episodic memory tests. A large number of recent studies have examined the effects of elaboration on implicit memory tests. Implicit memory (Schacter 1987) demonstrates an influence of a prior episode that is independent of the participant’s conscious awareness of that episode. Implicit memory tests generally attempt to show that performance on a task is enhanced by a prior episode even though participants are not necessarily consciously aware of the effect of the prior episode on their performance. Examples of such tests include perceptual identification and fragment completion. In perceptual identification, words are presented very briefly (e.g., for 33 ms) and participants are asked to identify them. Some of the items presented in the perceptual identification test are items that were presented in an earlier study episode and the test attempts to see if identification is better for these old items than for new items. Differences in perceptual identification between old and new items are generally referred to as ‘priming.’ The fragment completion test operates analogously to the perceptual identification test, except that, instead of rapid presentation, letters are deleted from the to-be-identified item.

Graf and Mandler (1984) hypothesized that elaborative processes affect explicit memory tests such as recognition and recall, but not implicit memory tests such as perceptual identification and fragment completion. Their rationale was that implicit memory tests are primarily sensitive to integrative processing of perceptual information. Assuming that elaborative processing occurs following item perception, one would not expect variations in elaborative processing to influence the integration of perceptual information.

The literature investigating these hypotheses has generated a complex pattern of results, with some manipulations of elaborative processing (e.g., levels-of-processing, Jacoby and Dallas 1981) producing no detectable effects on implicit memory and other manipulations of elaborative processing (e.g., generation, Masson and MacLeod 1992) demonstrating robust effects on implicit memory tests. Further, even in the cases where effects were not detectable in individual experiments, meta-analyses have demonstrated effects. For example, Jacoby and Dallas (1981) demonstrated that levels-of-processing produced no effect on perceptual identification. However, Brown and Mitchell (1994) demonstrated effects of levels-of processing on perceptual identification in a meta-analysis. A final complication in this literature is that results demonstrating effects of elaborative processing on implicit memory tests may arise because explicit memory processes contaminate performance in the implicit memory test.

While the preceding summary demonstrates a literature in ferment, and one that has far from definitive conclusions, two hypotheses may help rationalize this complex pattern of results. First, it may be the case that the small effects demonstrated in meta-analyses arise from contamination by explicit memory processes. If this is the case, results such as those of Jacoby and Dallas (1981) support Graf and Mandler’s original hypothesis that elaborative processing does not necessarily influence performance on implicit memory tests. Second, there may be some heterogeneity in the effects of different types of elaborative processing on implicit memory tests.

Thus, while the levels-of-processing manipulation used in the experiment of Jacoby and Dallas (1981) may not affect performance on implicit memory tests, the generation manipulation used by Masson and MacLeod (1992) may do so. Establishing such heterogeneity in types of elaborative processes would be consistent with cognitive psychology’s tradition of differentiating types of encoding processes (see introductory sections).

6. Conclusion

The modern construct of elaboration arose from cognitive psychology’s emphasis on intentional processing and information transformation—an emphasis reflected in filter models of attention and stage models of memory. Its specific genesis was in attempts to explain the congruity effect demonstrated in the levels-of-processing literature. Elaboration of the relation between an orienting question and to-be-6recalled information was assumed to enhance performance, producing the congruity effect. Theoretical accounts of effects of elaboration have emphasized that the availability and distinctiveness of retrieval cues are critical to memory performance. The construct of elaboration has been widely applied with differences in the ability to create or use elaborations being proposed as explanations of differences in memory due to the effects of age, materials, participant’s knowledge, and type of memory test.

Bibliography:

1. Atkinson R C, Shiffrin R M 1968 Human memory: A proposed system and its control processes. In: Spence K W, Spence J T (eds.) The Psychology of Learning and Motivation: Advances in Theory and Research, 2. Academic Press, New York
2. Baddeley A D, Hitch G 1974 Working memory. In: Bower G H (ed.) The Psychology of Learning and Motivation, 8. Academic Press, New York
3. Barnes M A, Dennis M, Haefele-Kalvaitis J 1996 The effects of knowledge availability and knowledge accessibility on coherence and elaborative inferencing in children from six to fifteen years of age. Journal of Experimental Child Psychology 61: 216–41
4. Bower G H 1970 Organizational factors in memory. Cognitive Psychology 1: 18–46
5. Broadbent D E 1958 Perception and Communication. Pergamon, Oxford, UK
6. Brown A, Mitchell D 1994 A reevaluation of semantic vs. nonsemantic processing in implicit memory. Memory and Cognition 22: 533–41
7. Chase W G, Simon H A 1973 The mind’s eye in chess. In: Chase W G (ed.) Visual Information Processing. Academic Press, New York
8. Cherry K E, Park D C, Frieske D A, Rowley R L 1993 The effect of verbal elaborations on memory in young and older adults. Memory and Cognition 21: 725–38
9. Craik F I M, Jacoby L L 1979 Elaboration and distinctiveness in episodic memory. In: Nilsson L (ed.) Perspectives on Memory Research: Essays in Honor of Uppsala University’s 500^th Erlbaum, Hillsdale, NJ
10. Craik F I M, Lockhart R S 1972 Levels of processing: A framework for memory research. Journal of Verbal Learning and Verbal Behavior 11: 671–84
11. Craik F I M, Tulving E 1975 Depth of processing and the retention of words in episodic memory. Journal of Experimental Psychology 104: 268–94
12. Craik F I M, Watkins M 1973 The role of rehearsal in short term memory. Journal of Verbal Learning and Verbal Behavior 12: 599–607
13. Erdfelder E, Bredenkamp J 1998 Recognition of script-typical versus script-atypical information: Effects of cognitive elaboration. Memory and Cognition 26: 922–38
14. Graf P, Mandler G 1984 Activation makes words more accessible, but not necessarily more retrievable. Journal of Verbal Learning and Verbal Behavior 23: 553–68
15. Horiba Y 1996 The role of elaborations in L2 text memory: The effects of encoding task on recall of casually related sentences. Modern Language Journal 80: 151–64
16. Hunt R R, Elliott J M 1980 The role of nonsemantic information in memory: Orthographic distinctiveness effects upon retention. Journal of Experimental Psychology: General 109: 49–74
17. Jacoby L L, Dallas M 1981 On the relationship between autobiographical memory and perceptual learning. Journal of Experimental Psychology: General 110: 306–40
18. Kintsch W, van Dijk T A 1978 Toward a model of text comprehension and production. Psychological Review 85: 363–94
19. Lea R B 1995 On-line evidence for elaborative logical inferences in text. Journal of Experimental Psychology: Learning Memory and Cognition 21: 1469–82
20. Masson M E J, MacLeod C M 1992 Reenacting the route to interpretation: Enhanced preceptual identification without prior perception. Journal of Experimental Psychology: General 121: 145–76
21. Paivio A 1971 Imagery and Verbal Processes. Holt, New York
22. Raaijmakers J G W, Shiffrin R M 1981 Search of associative memory. Psychological Review 88: 93–134
23. Reder L M, Anderson J R, Bjork R A 1974 A semantic interpretation of encoding specificity. Journal of Experimental Psychology 102: 648–56
24. Schacter D L 1987 Implicit memory: History and current status. Journal of Experimental Psychology: Learning Memory and Cognition 13: 501–18
25. Scruggs T E, Mastropieri M A, Sullivan G S 1994 Promoting relational thinking: Elaborative interrogation for students with mild disabilities. Exceptional Children 60: 450–57
26. Willoughby T, Waller G T, Wood E, MacKinnon G E 1993 The effect of prior knowledge on an immediate and delayed associative learning task following elaborative interrogation. Contemporary Educational Psychology 18: 36–46
27. Woodward A, Bjork R, Jongeward R 1973 Recall and recognition as a function of primary rehearsal. Journal of Verbal Learning and Verbal Behavior 12: 608–17
28. Zabrucky K, Moore D 1995 Elaborations in adults’ text recall: Relations to working memory and text recall. Experimental Aging Research 21: 143–58