Psychology Of Expert Memory Research Paper

In the remainder of the paper, some of the fundamental phenomena that have been discovered about human expertise will be outlined. An argument will be advanced that the Development of expertise is largely a matter of reorganizing knowledge and cognitive processes to do tasks efficiently and effectively. Experts restructure their inner workings—knowledge and procedures—for efficient application to their work-a-day environments (see also Ericsson and Lehmann 1996).

1. Larger Cognitive Units

As people gain increasing experience working in an environment, they cognitively organize the environment into larger units. This is the classic and one of the best-established phenomena of expertise, supported by a long line of studies that starts with the game of chess.

In the 1960s and early 1970s, De Groot (1965) and Chase and Simon (1973) studied master level and less expert chess players. In the basic task from these studies, subjects were shown a chess board with pieces representing game states from real games. Subjects were shown the boards for only five seconds and were then asked to reproduce the board they had seen. Experts, after a brief glance, were able to reproduce much more of the configuration than novices. In the Simon and Chase studies, expert advantage was about 4–1. In the De Groot studies, expert performance was nearly perfect (for 25 piece boards), while novices were able to reproduce about five items—or about the number of items that can be held in short-term memory. In related studies, the same investigators showed that after presentation of a board, experts were able to choose better next moves than novices.

The general explanation for expert superiority in the chess studies involves ‘chunking’ in perception and memory. With experience, experts acquire a large ‘vocabulary’ or memory store of board patterns involving groups of pieces or ‘chunks.’ A chunk is a perceptual or memory structure that bonds a number of more elementary units (e.g., chess pieces) into a larger organization. When experts see a chess setup from a real game, they are able to recognize these familiar patterns. They can then associate these patterns with moves stored in memory that have proved effective in the past. Novices have not been exposed to game configurations enough to have developed these kinds of patterns. Hence they deal with the board in a piece-by-piece manner. Similarly, when experts are presented with chess boards containing randomly placed pieces that do not enable the experts to take advantage of established patterns, they perform at the same level as novices.

These basic chess phenomena have been replicated many times—in chess, but also in numerous other fields (e.g., the games of bridge, Engle and Bukstel 1978; GO, Reitman 1976; and electronics, Egan and Schwartz 1979). In most studies of this sort, it is really the chunk size that has been shown to be larger for experts. Both the novice and the expert are subject to the same limitations of short-term memory (7±2 units, Miller 1956). However, expert chunks are larger. Hence, whereas a chess novice sees a number of isolated chess pieces, the expert recognizes about the same number of larger units—for example, one might be a ‘king defense configuration’—each composed of a number of chess pieces.

The fact that expert organizational units are larger is interesting; it indicates an expert adaptation useful for expanding the functional size of short-term or working memory. However, it is not the general finding that is most interesting from these kinds of studies. What is perhaps more revealing is the nature and/organization of the ‘chunks’ or memory structures experts use, discussed next.

2. Abstracted, Functional Organization

Some studies, using variants of the standard Simon and Chase chessboard paradigm, have examined the actual nature of expert and novice ‘chunks’ or knowledge structures. Table 1 presents a general overview of the differences that emerge.

Chase and Simon (1973) examined the nature of chunks experts used in reconstructing a briefly presented board. Expert chunks were found largely to comprise attack and defense configurations, tied to more general strategic aspects of the game. It is not clear how novice units were organized.

Studies of bridge (Charness 1979, Engle and Bukstel 1978) and electronics (Egan and Schwartz 1979), following the Chase and Simon paradigm, showed similar results among experts. In the bridge studies, experts and novices were presented depictions of four- handed bridge deals and were required to reproduce these deals. Experts reproduced the cards by suit across hands, remembering cards of the same suit from three hands and inferring the fourth, an organization useful in playing the game of bridge. Novices recalled the cards by order of card rank within hands, perhaps a carryover from other card games, for example, poker. In the electronics studies, subjects were shown an electronic circuit diagram, which they were then requested to reproduce. Experts grouped individual diagram components into major electronic components (e.g., amplifiers, filters, rectifiers), while novice organization was largely based on the spatial proximity of basic diagram symbols.

Similar results have been shown by studies from other fields, using somewhat different methodologies. Voss and co-workers (Spilich et al. 1979) studied baseball aficionados and more casual baseball ob- servers. Subjects were presented a colorful description of a half-inning of baseball and were then to recall the inning. Expert recall was structured by major goal- related sequences of the game such as advancing runners, scoring runs, and preventing scoring. Novice recall involved less integral components, for example, the weather and the crowd mood, and did not capture goal-directed, sequential activity as well.

Two studies of computer programming are also noteworthy. In a study of expert and novice ALGOL programmers, McKeithen et al. (1981) presented a list of 21 commands in the ALGOL language to ALGOL experts, students after one ALGOL course, and students at the beginning of an ALGOL course. Subjects were given 25 recall trials after initial learning of the list. Recall organization by pre-ALGOL students included organization by surface features of commands (e.g., commands with the same beginning letter or same length of command name), and groups of commands forming natural language segments (e.g., STRING IS NULL BITS) which have no conceptual meaning within ALGOL. Experts, however, grouped commands that formed mini ALGOL algorithms (e.g., formation of loops) or represented types of ALGOL data structures. Students after an ALGOL course produced groupings that were a mixture of surface related and conceptual ALGOL organizations. A similar study (Adelson 1981) presented a list containing three intact computer programs, scrambled together and out of order, to expert and novice programmers who were required to recall the list. Over recall trials, experts reconstructed the original three major algorithms, while novice recall organization was according to syntactic similarities in program statements, regardless of which program they had come from.

Other findings have come from work in physics (Chi et al. 1981) and medicine (Feltovich et al. 1984). In the basic task from the physics study, problems from chapters in an introductory physics text were sampled and placed on individual cards. Expert (professors and advanced graduate students) and novice (college students after their first mechanics course) physics problem solvers sorted problems into groups they ‘would solve in a similar manner.’ The finding was that experts created groups based on the major physics laws (conservation and force laws) applicable in solution. Novice groupings were organized by salient objects (e.g., springs, inclined planes) and features contained in the problem statement itself. Similarly, in studies of expert and novice diagnoses within a subspecialty of medicine, it was found that expert diagnosticians organized diagnostic hypotheses according to the major physiological issue relevant in a case, while novice hypotheses were more isolated and seemed more dependent on particular patient cues.

Zeitz (1997) has reviewed these and more recent studies of this type, investigating what she calls experts’ use of ‘Moderately Abstracted Conceptual Representations’ (MACRs), representational abstractions of the type we have just discussed. She proposes numerous ways that such abstraction aids the efficient utilization of knowledge and reasoning by experts. These include: (a) the role of abstracted representations in retrieving appropriate material from memory (e.g., Chi et al. 1981); (b) the schematic nature of MACRs in integrating information, suggesting what information is important, activating appropriate prior knowledge, providing guidance for appropriate action, and providing support argumentation for a line of approach (e.g., Phelps and Shanteau 1978, Schmidt et al. 1989, Voss et al. 1983); (c) aiding productive analogical reasoning (e.g., Gentner 1988); and (d) providing guidance for the elimination of unimportant information and unnecessary detail (e.g., Patel and Groen 1991).

In summary, research conducted in the last quarter of the twentieth century indicates that expert knowledge organization develops in the direction of in-creased value for doing the work of the expert’s field. Novices also impose organization—but it is less useful—for example, organization by items named in a situation, proximity of entities to others, or superficial analogy.

3. Automaticity

Practice is almost a defining property of expertise. Experts are generally individuals with extreme amounts of practice on a circumscribed set of tasks of a work environment. For example, some expert radiologists who have been studied estimated that they have analyzed more than half a million radiographs (X-rays) in their careers (Lesgold et al. 1988).

A general theme of research on automaticity is that the character of cognitive operations changes with practice. Operations that are initially slow, serial, and demand conscious attention become fast, less deliberate, and can run in parallel with other processes (Schneider and Shiffrin 1977). With enough practice, one can do several things at once.

There are many examples in the literature of the effects of practice on automaticity. For example, expert typists can type and recite nursery rhymes at the same time (Shaffer 1975). Skilled abacus operators can answer routine questions (‘What is your favorite food?’) without loss of accuracy or speed in working with the abacus (Hatano et al. 1977). After six weeks of practice (one hour per day), college students could read unfamiliar text while simultaneously copying words read by an experimenter—without decrement in reading speed or comprehension (Spelke et al. 1976).

Automaticity is important to expertise. In complex skills with many different cognitive components, some of the more basic ones (e.g., fundamental decoding encoding of input) must be automated if higher levels skills such as reasoning, inference, and integration are ever to be proficient (e.g., Lesgold and Resnick 1982). There is also a possible interaction between automaticity of processes and the usability of available knowledge. Investigators (e.g., Feltovich et al. 1984, Jeffries et al. 1981) have suggested that a major limitation of novices is in their inability, situationally, to utilize knowledge that they in some sense possess. Problems in knowledge usability may be associated with overload or inefficiency in using working (or short-term) memory. The usable knowledge of experts, in turn, may devolve from the subordination of many task components to automatic processing, which increases capacity for more controlled management of memory and knowledge application (cf. Perfetti and Lesgold 1979).

Although many years of experience are a necessary requirement for the development of expertise, extensive experience alone is not sufficient. The seminal work of Ericsson indicates that long-term practice needs to be thoughtful, reflective, and self-critical; that is, it must constitute ongoing ‘deliberate practice’ (e.g., Ericsson and Charness 1994, Ericsson and Lehmann 1996). It is not sufficient just to ‘go through the motions.’

4. More Recent Developments

Recent research on expertise has examined the contexts in which expertise develops and is practiced. An impetus for this research is to understand how the contexts of expertise provide the support (including that supplied by other talented co-workers and sponsors) and opportunities (e.g., access to the unusual, tricky cases in one’s field) that enable an individual’s development toward expertise to progress (e.g., Agnew et al. 1997, Stein 1997). Another area of investigation involves how the nature of expertise varies with changing contexts, including the nature of expertise when work environments change so frequently that it is impossible to practice any one, circumscribed set of skills over a long period of time (Feltovich et al. 1997a, 1997b, Hoffman et al. 1997, Prietula et al. 2000).

5. Conclusions

A general argument has been made that expert knowledge structures and procedures are reorganized in directions that enable effective application to task demands of a working environment. Most of these changes are adaptations that enable utilization of large amounts of information in the context of limited internal processing resources—in particular those imposed by the small capacity of short-term or working memory.

Grouping or chunking on information structures and procedure components functionally increases the size of working memory and its efficiency. More information can be considered for each ‘unit’ of working memory. Selectivity, discrimination, and abstraction insure that only the most useful information is thrown into competition for resources. Automaticity is a means of restructuring some procedures so that working memory is largely circumvented—freeing resources for other cognitive chores. It is a tension between high information flow and limited internal resources that encourages the development of strategies for the efficient use of information and knowledge. Expertise is largely a progressive adaptation by a limited cognitive system to the great demands posed by important, complex cognitive tasks.

Bibliography:

1. Adelson B 1981 Problem solving and the development of abstract categories in programming languages. Memory and Cognition 9: 422–33
2. Agnew N M, Ford K M, Hayes P J 1997 Expertise in context: Personally constructed, socially selected, and reality-relevant. In: Feltovich P J, Ford K M, Hoffman R R (eds.) Expertise in Context: Human and Machine. AAAI MIT Press, Menlo Park, CA, pp. 219–44
3. Bransford J D, Stein B S, Shelton T S, Owings R A 1981 Cognition and adaptation: The importance of learning to learn. In: Harvey J H (ed.) Cognition, Social Behavior and the Environment. Lawrence Erlbaum, Hillsdale, NJ
4. Charness N 1979 Components of skill in bridge. Canadian Journal of Psychology 33: 1–16
5. Chase W G, Simon H A 1973 Perception in chess. Cognitive Psychology 4: 55–81
6. Chi M T H, Feltovich P J, Glaser R 1981 Categorization and representation of physics problems by experts and novices. Cognitive Science 5: 121–52
7. Chi M T H, Glaser R, Farr M J 1988 The Nature of Expertise. Lawrence Erlbaum, Hillsdale, NJ
8. DeGroot A D 1965 Thought and Choice in Chess. Basic Books, New York
9. Egan D E, Schwartz B J 1979 Chunking in recall of symbolic drawings. Memory and Cognition 7: 149–58
10. Engle R W, Bukstel L 1978 Memory processes among bridge players of differing expertise. American Journal of Psychology 91: 673–89
11. Ericsson K A 1996 The Road to Excellence: The Acquisition of Expert Performance in Arts and Sciences, Sports and Games. Lawrence Erlbaum, Mahwah, NJ
12. Ericsson K A, Charness N 1994 Expert performance: Its structure and acquisition. American Psychologist 49: 725–47
13. Ericsson K A, Lehmann A C 1996 Expert and exceptional performance: Evidence on maximal adaptations on task constraints. Annual Review of Psychology 47: 273–305
14. Ericsson K A, Smith J 1991 Toward a General Theory of Expertise: Prospects and Limits. Cambridge University Press, Cambridge, UK
15. Feltovich P J, Ford K M, Hoffman R R 1997a Expertise in Context: Human and Machine. AAAI MIT Press, Menlo Park, CA
16. Feltovich P J, Johnson P E, Moller J H, Swanson D B 1984 LCS: The role and development of medical knowledge of diagnostic expertise. In: Shortliffe E H, Clancey W J (eds.) Readings in Medical Artificial Intelligence: The First Decade. Addison-Wesley, Reading, MA, pp. 275–319
17. Feltovich P J, Spiro R J, Coulson R L 1997b Issues of expert flexibility in contexts characterized by complexity and change. In: Feltovich P J, Ford K M, Hoffman R R (eds.) Expertise in Context: Human and Machine. AAAI MIT Press, Menlo Park, CA, pp. 125–46
18. Gentner D 1988 Metaphor as structure mapping: The relational shift. Child Development 59: 47–59
19. Hatano G, Miyake Y, Binks M B 1977 Performance of expert abacus operators. Cognition 5: 57–71
20. Hoffman R R 1992 The Psychology of Expertise: Cognitive Research and Empirical AI. Springer-Verlag, New York
21. Hoffman R R, Feltovich P J, Ford K M 1997 A general framework for conceiving of expertise and expert systems in context. In: Feltovich P J, Ford K M, Hoffman R R (eds.) Expertise in Context: Human and Machine. AAAI MIT Press, Menlo Park, CA, pp. 543–80
22. Jeffries R, Turner A A, Polson P G, Atwood M E 1981 The processes involved in software design. In: Anderson J R (ed.) Cognitive Skills and Their Acquisition. Lawrence Erlbaum, Hillsdale, NJ
23. Lesgold A M, Resnick L B 1982 How reading difficulties develop: Perceptives from a longitudinal study. In: Das J P, Mulcahy R, Walls A E (eds.) Theory and Research in Learning and Learning Disabilities. Plenum, New York
24. Lesgold A M, Rubinson H, Feltovich P J, Glaser R, Klopfer D, Wang Y 1988 Expertise in a complex skill: Diagnosing x-ray pictures. In: Chi M T H, Glaser R, Farr M J (eds.) The Nature of Expertise. Lawrence Erlbaum, Hillsdale, NJ, pp. 311–42
25. McKeithen K B, Reitman J S, Reuter H H, Hirtle S C 1981 Knowledge organization and skill differences in computer programmers. Cognitive Psychology 13: 307–25
26. Miller G A 1956 The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review 63: 81–97
27. Newell A, Simon H A 1972 Human Problem Solving. Prentice Hall, Englewood Cliffs, NJ
28. Patel V L, Groen G J 1991 The general and specific nature of medical expertise. In: Ericsson K A, Smith J (eds.) Toward a General Theory of Expertise: Prospects and Limits. Cambridge University Press, Cambridge, UK, pp. 93–125
29. Perfetti C A, Lesgold A M 1979 Coding and comprehension in skilled reading. In: Resnick L B, Weaver P (eds.) Theory and Practice of Early Reading. Lawrence Erlbaum, Hillsdale, NJ
30. Phelps R H, Shanteau J 1978 Livestock judges: How much information can an expert use? Organizational Behavior and Human Performance 21: 209–19
31. Prietula M J, Feltovich P J, Marchak F 2000 Factors influencing analysis of complex cognitive skills: A framework and example from industrial process control. Human Factors 42: 56–74
32. Reitman J 1976 Skilled perception in GO: Deducing memory structure from inter-response times. Cognitive Psychology 8: 336–56
33. Schmidt J A, McLaughlin J P, Leighton P 1989 Novice strategies for understanding paintings. Applied Cognitive Psychology 3: 65–72
34. Schneider W, Shiffrin R M 1977 Controlled and automatic human information processing: 1. Detection, search and attention. Psychological Review 84: 1–66
35. Shaffer L H 1975 Multiple attention in continuous verbal tasks. In: Rabbitt P M A, Dornic S (eds.) Attention and Performance V. Academic Press, New York
36. Spelke E, Hirst W, Neisser U 1976 Skills of divided attention. Cognition 4: 215–30
37. Spilich G J, Vesonder G T, Chiesi H L, Voss J F 1979 Text processing of domain-related information for individuals with high and low domain knowledge. Journal of Verbal Learning and Verbal Behavior 18: 275–90
38. Stein E W 1997 A look at expertise from a social perspective. In: Feltovich P J, Ford K M, Hoffman R R (eds.) Expertise in Context: Human and Machine. AAAI MIT Press, Menlo Park, CA, pp. 181–94
39. Voss J F, Greene T R, Post T A, Penner B C 1983 Problem solving skill in the social sciences. In: Bower G (ed.) The Psychology of Learning and Motivation. Academic Press, New York, Vol. 17, pp. 165–213
40. Zeitz C M 1997 Some concrete advantages of abstraction: How experts’ representations facilitate reasoning. In: Feltovich P J, Ford K M, Hoffman R R (eds.) Expertise in Context: Human and Machine. AAAI MIT Press, Menlo Park CA, pp. 43–65