Novex Analysis: A Cognitive Science Approach to Instructional Design

Novex Analysis:
A Cognitive Science Approach To Instructional Design
James C Taylor
James C Taylor, Vernon White Chair of Distance Education, is Professor and Director of the Distance Education Centre,
University of Southern Queensland, Australia.
In a recent review of the increasing influence of cognitive psychology on instructional design, Tennyson (1992) proposed a cognitive system "to serve as an initial learning paradigm for the field of instructional design theory" (p.40),and invited others to extend the approach. The present paper extends Tennyson‘s cognitive learning theory by: (i) elaborating a model of the human information processing system; (ii) outlining an approach to instructional design based on the analysis of the cognitive structures of both novices and experts ("Novex Analysis"); (iii) exemplifying a critical feature of the approach by delineating key elements of the knowledge base of an expert in neuroscience; and (iv) providing basic guidelines for evaluating the efficacy of the proposed Novex analysis approach to instructional design. In essence, the approach recognizes the need for the cognitive modelling of "novice systems" as well as "expert systems" if the efficacy of instructional design is to be enhanced. Novex analysis demands a team approach to design, which not only involves the instructional designer and content expert, but also a representative sample of relative novices drawn from the student target population. The major focus of this paper, however, is to outline the cognitive science underpinnings of the approach.
A Model of the Human Information Processing System
The "Dimensions of Processing Model" of human information processing (Taylor, 1983; Taylor & Evans, 1985) is consistent with the proposal by Glaser (1982) that the psychology of instruction should attempt to understand the development of cognitive structures and processes that characterize the proficient performance of cognitive skills by experts in particular subject matter domains. A fundamental issue for such an approach is the availability of a theoretical framework which is capable of representing the acquisition, structure and organisation of the knowledge base underlying expert cognitive skill performance. The Dimensions of Processing Model was developed for just such a purpose.
The generation of a comprehensive model of human information processing is probably somewhat ambitious at a time when cognitive science is characterised by extensive theory construction and related empirical work aimed at investigating the fundamental mechanisms underlying human cognition. In such a complex field of endeavour, it is inevitable that different workers approach issues from different directions and yield different levels of explanation of essentially the same phenomenon. This attempt to generate a unifying framework must therefore be viewed as an effort to work at a level of description which is not encumbered by excessive detail, but which is sufficiently explicit to engender the formulation of hypothesised memory structures and associated instructional design guidelines that are empirically testable.
The structure and organisation of the knowledge base underlying expert cognitive skill performance should reflect the overall structure of an explicit model of memory. Indeed,a model of memory is fundamental to any model of human information processing. Most cognitive psychologists distinguish at least three kinds of memory: a sensory register under executive control, a short-term or working memory and a long-term memory (Bovy, 1981; Bruning, 1983; Frederiksen, 1984; Tennyson, 1992). The Dimensions of Processing Model incorporates three interrelated major elements: executive memory (Gagne, 1977), working memory (Baddeley, 1981) and long-term memory (Kumar, 1971), all of which attempt to account for observed processes underlying human information processing.
Consistent with the focus on the structure and organisation of knowledge underlying expert cognitive skill performance, the major emphasis within the model is on long-term memory. At the same time, however, the framework represents an attempt to integrate the relatively static conceptions of memory structures with the more dynamic conceptions underlying research on control and processing mechanisms, thereby providing a perspective for understanding the human information processing system as a whole. Further, the model embodies the synthesis of a number of conceptual orientations, including relatively contemporary emphases on item-specific and relational information (Hunt & Einstein, 1981) schema abstraction (Elio & Anderson, 1981), strategic knowledge (Chi, 1978; Greeno, 1978), and metacognition (Flavell, 1979, 1985; Schmitt & Newby, 1985; Garner & Alexander, 1989). This synthesis rests on the assumption that the nature of knowledge structures in long-term memory depends on the type of information processing whereby these cognitive structures were generated. An elaborated overview of the Dimensions of Processing Model is presented in Figure 1.
Figure 1
The Dimensions of Processing Model

Domain-specific Knowledge
The most significant departure of the Dimensions of Processing Model from the conceptual heuristic proposed by Tennyson (1992) is that in the present model, long-term memory is conceived as the repository for schemas associated with domain-specific knowledge, while executive memory is concerned with domain-independent cognition and associated control processes, including contextual knowledge ("knowing why, when and where" to deploy specific concepts, rules and principles). In essence, executive memory is conceived as managing the deployment of control processes (eg perception, attention, resources) and general cognitive strategies, which may be applied to any domain of knowledge. In contrast, long-term memory is regarded as being concerned solely with the content and organisation of memory structures and processes that are primarily a function of domain-specific knowledge, sometimes referred to as discipline-specific or content-specific knowledge (Alexander, Schallert & Hare, 1991). Such knowledge was described by Popper (1972) as "objective knowledge", which is purported to exist independently of human information processing. Popper argued for the independent existence of "the objective contents of thought" (1972, p.106), which he equated with theoretical systems, problem situations, critical arguments and the contents of journals and books. From an education and training perspective, many instructional programs could be regarded as aiming to embed such domain-specific objective knowledge in the minds of students.
The approach adopted here is to consider cognitive structure from a conceptual heuristic standpoint, which differentiates memory structures and processes that are primarily a function of the structure of domain-specific objective knowledge from other features of human cognitive processing. Within the context of the Dimensions of Processing Model, executive memory, working memory and their associated componential processes are considered to be inherent features of the human information processing system. These structures and processes can be usefully regarded as basic properties of all cognitive functioning, whereas the following elements of cognitive structure are primarily a function of domain-specific objective knowledge: item-specific, relational and strategic knowledge. Objective knowledge is also distinguished from affective processing and empirical processing, which represent more subjective states of consciousness, resulting from potentially idiosyncratic responses to personal experience. By their very nature, both objective knowledge and subjective knowledge are domain specific. As such, they are significant variables in the design of instruction, and are therefore worthy of furher consideration.
Item-specific and Relational Knowledge
Two of the three dimensions of objective knowledge, item-specific and relational knowledge respectively, were derived primarily from the work of Hunt and Einstein (1981), who highlighted the distinction between item-specific and relational information. The former results from processing characterized by the encoding of distinctive information unique to each separate input event, whereas relational information is generated by processing involving the abstraction of common features shared by elements or events at input. Hunt and Einstein produced evidence that optimal memory performance depended on the existence of a memory trace that resulted from the performance of a task which entailed both item-specific and relational processing. Similarly, Elio and Anderson (1981) provided evidence that supported a schema abstraction model in which transfer is a function of similarity both to specific category instances and to higher order category information abstracted from those instances. This basic notion of abstracted schematic knowledge has been extended to incorporate the organisation of the knowledge base, which in the case of the expert entails a potentially fluid, kaleidescopic array of cognitive structures underpinning a wide range of relationships between domain-specific ideas.
The distinction between item-specific and relational knowledge is consistent with Glaser‘s (1984; 1991) interest in the differences between the content and organisation of the knowledge bases of experts and novices. The former is typically tightly organised in hierarchical fashion, whereas the knowledge of the novice is usually characterized by a series of loosely related ideas. Similarly, the way in which knowledge is organised is a central theme of research on chunking which uncovered hierarchical knowledge structures (Akin,1980; Larkin & Reif, 1979). Indeed, some authors have argued that hierarchical structure is fundamental to cognition (Chi & Rees, 1983; Miller, Galanter Pribram, 1960). The notion of relational knowledge appears to provide an appropriate vehicle for the exploration of the organisation of the knowledge bases of both experts and novices.
Strategic Knowledge
The postulated third dimension of domain-specific objective knowledge results from strategic processing. Strategic knowledge is essentially concerned with applying domain-specific knowledge to domain-specific tasks (Chi, 1978; Greeno, 1978). It should not be confused with general strategic processes (eg means-ends analysis), which are part of executive memory, and which are activated when a person is confronted by a task environment for which there is no domain-specific knowledge available. The operation of domain-specific strategic processes is consistent with Landa‘s (1974, 1976, 1993) algo-heuristic theory of instruction, which is based on the explicit teaching of component cognitive operations underlying task performance, problem solving or decision making in specific knowledge domains. Strategic knowledge resulting from task-specific processing within a particular knowledge domain is the postulated third dimension of a memory trace, which also incorporates item-specific and relational knowledge. The remaining two dimensions are the affective, concerned with the emotional-motivational aspects of cognition, and empirical processing, which highlights the influence of authentic experience on human memory. These latter two dimensions represent subjective knowledge, reflecting individual responses to experiences associated with particular knowledge domains.
Affective Knowledge
As Tennyson (1992) pointed out, affective knowledge "needs to be considered during the acquisition of knowledge and as part of the knowledge base for content" (p.40). Although a promising beginning was made by Bower (1981) in proposing the existence of an emotional node, and while Keller (1983) has developed a motivational-design model of instruction, there is considerable scope for further work on documenting the affective aspects of the knowledge bases of experts. In particular, the core values that are typically held in common by members of a profession, and associated codes of professional practice, need to be incorporated in the specification of expert knowledge bases. It seems reasonable to suggest that any comprehensive description of an expert knowledge base needs to include an affective dimension, which is treated as being thoroughly integrated with the other dimensions.
Empirical Knowledge
The empirical dimension completes the proposed model of knowledge structures in long-term memory. In the first place, empirical knowledge is the same as episodic knowledge (Tulving, 1972) in the sense that it keeps a record of experience in a historico-spatial framework. This aspect of empirical knowledge is not especially significant when viewed from a perspective aimed at elucidating the acquisition and performance of cognitive skills. It is important, however, in that it subsumes the psychomotor skill component of cognitive structure resulting from actual skill performance. The inclusion of a psychomotor dimension under the auspices of empirical knowledge is in the interests of parsimony, since a psychomotor trace can result only from actual task performance, which is, of course, a function of experience.
The impact of empirical processing on cognitive structure depends on the quality and intensity of experience, specifically: the extent to which the experience was personal or vicarious, and the extent to which events were authentic, simulated or artificial. Consider, for example, the difference in the affective component of the cognitive structure of a surgeon, who has completed his or her first major operation, with that of neophytes observing in theatre but yet to make their first incision, despite their having an equivalent thorough grounding in relevant theory. Similarly, the authenticity and intensity of first-hand experience seems likely to stimulate the refinement and strengthening of those dimensions of cognitive structure (item-specific, relational and strategic) emanating from the structure of objective knowledge. Further, the skill required by the surgeon in the operating theatre highlights the inter-relationships between the cognitive, affective and psychomotor domains, encapsulated in the "head, heart and hand" analogy first proposed by Lippert and Farmer (1984, pp.11-13). In the sense that empirical processing usually entails simultaneous processing in all three domains, it acts as a linking mechanism providing the adhesive that binds domain-specific cognitive structures together. More importantly, though, it emphasises the need for students to practice cognitive skills in realistic environments.
Target Subject Matter
The provision of the opportunity for students to practice cognitive skills in authentic working environments, although desirable, is beyond the scope of many education and training programs. The importance of exposure to first-hand experience in realistic environments is, however, not only related to pragmatic considerations and the readiness of the student target population, but is also a function of the nature of the target subject matter. For example, Chang, Crombag, van der Drift and Moonen (1983) drew attention to the important distinction between symbolic and tangible reality (p.119). Some subject matter, such as mathematics, logic and language studies, consists entirely of what could be termed "symbolic reality". In such fields of study, students can presumably be exposed to an appropriate range of learning activities to ensure the development of a sufficiently comprehensive knowledge base. Where subject matter demands access to specific equipment (as for example in the teaching of science and engineering), particular environments (as in the teaching of land surveying or archaeology) or certain types of people (as in the teaching of counselling psychology or psychiatry), provision of appropriate access to such "tangible reality" (Chang, Crombag, van der Drift & Moonen, 1983) may be problematic. The extent to which direct experience of tangible reality is necessary for the generation of an expert knowledge base is a key consideration in the design of instruction. The need for first-hand experience of tangible reality may not be critical at all stages of a program. For example, in the early stages of skill acquisition it may be feasible to use simulations with opportunities for practice under authentic conditions being reserved for the refinement of skills already acquired at some level of mastery. Nevertheless, consideration of the extent to which various types of expert knowledge bases in different subject matter domains demand different degrees of direct experience (empirical processing) with tangible rather than symbolic reality is an important factor in making the decisions, which underpin the Novex analysis approach to instructional design.
Novex Analysis Approach to Instructional Design
The Novex analysis approach to instructional design entails the following nine steps:
(i) specify the domain-specific cognitive skill performance that is to be the terminal objective of instruction;
(ii) analyse the underlying declarative knowledge base of an expert (or experts) in the field of study primarily in terms of relational and strategic knowledge, but with due consideration also being given to affective and empirical knowledge;
(iii) use the framework provided by the expert knowledge base to measure the extant knowledge bases of the relative novices, who constitute the student target population for the proposed program of instruction;
(iv) design an advance organiser in light of salient elements of the knowledge bases of both novices and experts;
(v) present the structure and source of the expert knowledge base to the students through the gradual elaboration of a series of graphic organisers and associated strategic algorithms and/or heuristics to provide appropriate ideational scaffolding;
(vi) design a range of learning activities that demand different types of information processing (eg relational, strategic, affective, empirical) and provide associated model answers so that students can undertake performance self-evaluation;
(vii) apply the instructional treatment to the students;
(viii) provide performance related feedback to students through tutor marked assignments;
(ix) assess the levels of expertise achieved by students, thereby evaluating the efficacy of the instructional treatment to engender the novice to expert shift.
The Novex analysis approach to instructional design describes a process whereby the knowledge base of an expert is first analysed in terms of the structure of domain-specific objective knowledge. Specifically, the expert knowledge base is delineated according to the two major dimensions of objective knowledge, namely: relational and strategic knowledge. To provide a less abstract context for this discussion, key elements of the knowledge base of an expert in neuroscience (Bower & Taylor, 1992)are provided. This expert knowledge base is illustrated as a series of elaborations (Reigeluth & Stein, 1983) representing relational knowledge (Figure 2) and an algorithm (Landa, 1974; 1976; 1993) for the diagnosis of a particular illness exemplifying domain-specific strategic knowledge (Figure 3). The scope of the relational knowledge base is represented further in Figure 4.
Figure 2
An example of relational knowledge in neuroscience represented as a series of elaborations

Figure 3
An example of strategic knowledge in neuroscience represented as an algorithm

Figure 4
Overview of relational knowledge in neuroscience

Delineation of the expert knowledge base in terms of relational and strategic knowledge is the essential first step in the Novex analysis approach. Subsequent consideration should also be given to the affective and empirical knowledge components of the knowledge base. From an education and training perspective, the expert knowledge base provides a detailed enunciation of the terminal objective of an instructional program. In essence, the design of instruction is aimed at effecting the shift from novice to expert by creating a series of learning activities that will enable novices to construct and thereby replicate key elements of the organisation and content of the knowledge base of the expert in their own cognitive structure. Once the structure and content of the expert knowledge base have been delineated, the next step in the approach is to analyse the extant knowledge bases of the novices in the student target population in terms of this framework. The knowledge bases of the novices are likely to lack the coherence and connectivity of that of the expert, and may need to be represented as somewhat fragmented item-specific knowledge rather than organised frameworks of relational and strategic knowledge. Similarly, the affective and empirical dimensions of the knowledge bases of novices are unlikely to match the comprehensive richness of that of the expert. This assessment of the prior knowledge of novices relative to the explicit knowledge base of the expert completes the Novex analysis and provides a solid foundation for the next step in the approach: the design of an advance organiser (Ausubel,1968).
The design of an advance organiser can, of course, be based on the empirical data provided by the Novex analysis. This data provides the instructional designer with precise and explicit parameters for the design of an advance organiser, which is unlike other strategies such as concept mapping, in that "students are not ordinarily likely to be able to create them" (West, Farmer & Wolf, 1991, p.115). The next step entails the design of a number of graphic organisers, involving the gradual elaboration of the expert knowledge base (as illustrated in Figure 2). These graphic organisers provide an explicit ideational scaffolding for the generation of the expert knowledge base, which is the terminal objective of instruction. This knowledge base must, of course, be constructed by the student as a result of active participation in a series of carefully structured learning activities, each entailing a prescribed type of information processing according to the particular dimension of knowledge (eg relational, strategic, affective, empirical) to be generated.
It therefore behooves the instructional design team, a designer working in concert with at least one content expert and a representative sample of relative novices drawn from the student target population, to devise an appropriate range of learning activities within the framework provided by the detailed delineation of the expert knowledge base. Consistent with Anderson‘s (1982) proposition that there is a "need for an initial declarative encoding" (p.380), the recommended sequence of learning activities might well concentrate on the generation of relational knowledge prior to the generation of strategic knowledge and the subsequent activation of actual cognitive skill performance. Further, the emphasis in the Dimensions of Processing Model on the critical role of affective and empirical knowledge highlights the preference for using certain aspects of a situated learning approach to instruction, including learning activities based on complex issues in authentic contexts and the associated provision of scaffolding to enable novices to operate meaningfully in such realistic environments (Young, 1993). The degree to which graphic organisers are used to provide ideational scaffolding will, of course, vary according to the sophistication of the student target population. For instance, in the given example (Figures 2, 3 & 4) such scaffolding will be used sparingly, since the student target population consists of postgraduate medical students, who will be encouraged to generate their own algorithms and heuristics. Students will, of course, require feedback on such efforts.
Novex analysis endorses the use of self-assessment questions whereby students are provided with model answers and can therefore judge the efficacy of their own efforts. The use of self-evaluation as a cognitive strategy should enhance students‘ meta-cognitive skills, which in turn should facilitate learning. Such formative self-evaluation of performance will, of course, need to be supplemented by the provision of feedback through tutor marked assignments. This feedback should also be aimed at facilitating meta-cognitive awareness by assisting students to "plan, select, connect, tune, and monitor their own learning and transfer" (Clark, 1992, p.697). Similarly, the summative assessment of student achievement of cognitive skills is important since such data will be used partly to judge the efficacy of the instructional design decisions underlying the program.
Conclusion
Overall, the relative influence of subjective and objective knowledge over instructional design decisions will vary according to the nature of both the student target population and the target subject matter domain. Nevertheless, the essence of the instructional treatment is the design of learning activities that will deliberately induce different types of information processing (viz. relational, strategic, affective, empirical) aimed at generating the content and organisation of key elements of the expert‘s knowledge base in the minds of students. The detailed enunciation of the expert knowledge base in terms of taxonomic, often hierarchical subject matter structures and associated algo-heuristics, relevant components of affective knowledge and desirable first-hand experience in appropriate, realistic learning environments provides explicit parameters for making the necessary instructional design decisions.
In the final analysis, however, the efficacy of the actual design decisions is an empirical question. A recently published study (Taylor, 1991) provided empirical support for both the Dimensions of Processing Model and an associated instructional treatment devised to support an industrial training program in the field of water engineering. Further empirical investigations of the power of the Novex analysis approach are proceeding in the fields of contract law (Mason & Taylor, 1993), accountancy (Taylor & Thomas, in press) and decision support systems for sustainable dryland farming systems (Cox, Taylor & Thomas, 1993). Such investigations should lead to the refinement of the Novex analysis approach, which it is hoped can make a useful contribution to the knowledge bases of practising instructional designers.
References
Akin, O. Models of architectural knowledge. London: Pion, 1980.
Alexander, P.A., Schallert, D.L. and Hare, V.C. Coming to terms: How researchers in learning and literacy talk about knowledge. Review of Educational Research, 1991, 61, 3, 315-343.
Anderson, J.R. Acquisition of Cognitive Skill, Psychological Review, 1982, 89, 4, 369-406.
Anderson, J.R. Skill acquisition: Compilation of weak-method problem solutions. Psychological Review, 1987, 94, 192-210.
Ausubel, D.P. Education psychology: A cognitive view. New York: Holt, Rinehart & Winston, 1968.
Baddeley, A. The concept of working memory: A view of its current state and probable future development. Cognition, 1981, 10, 17-23.
Bovy, R.C. Successful instructional methods: A cognitive information processing approach. Educational Communication and Technology Journal, 1981, 29, 203-217.
Bower, A.J. and Taylor, J.C. A knowledge engineering approach to the teaching of neuroscience to medical students. Presentation to the Australasian Neuroscience Convention, Melbourne, February 1993.
Bower, G.H. Mood and memory. American Psychologist, 1981, 36, 129-148.
Bruning, I.L. An information processing approach to a theory of instruction. Educational Communication and Technology Journal, 1983, 31, 91-101.
Chang, T.M., Crombag, H.F., van der Drift, K.D.J.M., and Moonen, J.F. Distance Learning: On the design of an open university. Boston: Kluwer Nijhoff, 1983.
Chi, M.T.H. Knowledge structures and memory development. In R. Siegler (Ed.), Children‘s thinking; What develops?. Hillsdale, NJ: Erlbaum, 1978, 73-96.
Chi, M.T.H., and Rees, E.T. A learning framework for development. In M.T.H. Chi (Ed.), Contributions to human development: Trends in memory development research, Vol.9. Basel: Karger, 1983, 71-107.
Clark, R.E. How the cognitive sciences are shaping the profession. In H.D. Stolovitch and E.J. Keeps (Eds.), Handbook of Human Performance Technology. San Franciso: Jossey-Bass, 1992.
Cox, P., Taylor, J.C., and Thomas, N. Decision support systems for sustainable dryland farming. Paper presented to Workshop on Decision Support Systems, University of Southern Queensland, May 1993.
Elio, R., and Anderson, J.R.. The effects of category generalizations and instance similarity on schema abstraction. Journal of Experimental Psychology: Human Learning and Memory, 1981, 7, 397-417.
Flavell, J.H. Metacognition and cognitive monitoring: A new area of cognitive-developmental inquiry. American Psychologist, 1979, 34, 10, 906-911.
Flavell, J.H. Cognitive Development, 2nd edn. Englewood Cliffs, NJ: Prentice-Hall, 1985.
Frederiksen, N. Implications of cognitive theory for instruction in problem solving. Review of Educational Research, 1984, 54, 363-407.
Gagn? R.M. The Conditions of Learning, 3rd edn. New York: Holt, Rinehart and Winston, 1977.
Garner, R., and Alexander, P.A. Metacognition: Answered and unanswered questions. Educational Psychologist, 1989, 24, 2, 143-158.
Glaser, R. The maturing of the relationship between the science of learning and cognition and educational practice. Learning and Instruction, 1991, 1, 129-144.
Glaser, R. Instructional psychology: Past, present and future. American Psychologist, 1982, 37, 3, 292-305.
Glaser, R. Education and thinking: The role of knowledge. American Psychologist, 1984, 39, 93-109.
Greeno, J.G. Natures of problem-solving abilities. In W.K. Estes (Ed.), Handbook of Learning and Cognitive Processes, Vol.5. Hillsdale, NJ: Erlbaum, 1978.
Hunt, R.R., and Einstein, G.O. Relational and item-specific information in memory. Journal of Verbal Learning and Verbal Behaviour, 1981, 20, 497-514.
Keller, J.M. Motivational design of instruction. In C.M. Reigeluth (Ed.), Instructional-design theories and models: An overview of their current status. Hillsdale, NJ: Erlbaum, 1983, 383-434.
Kumar, V.K. The structure of human memory and some educational implications. Review of Educational Research, 1971, 41, 379-417.
Landa, L.N. Algorithmization in learning and instruction. Englewood Cliffs, NJ: Educational Technology Publications, 1974.
Landa, L.N. Instructional regulation and control: Cybernetics, algorithmization and heuristics in education. Englewood Cliffs, NJ: Educational Technology Publications, 1976.
Landa, L.N. Landamatics ten years later: An interview with Lev N. Landa. Educational Technology, 1993, 33, 6, 7-18.
Larkin, J.H., and Reif, F. Understanding and teaching problem solving in physics. European Journal of Science Education, 1979, 1, 191-203.
Lippert, G., and Farmer, J. Psychomotor skills in orthopaedic surgery. Baltimore: Williams and Wilkins, 1984.
Mason, R., and Taylor, J.C. A knowledge engineering approach to representing curriculum structure in legal education. The Law Teacher, 1993, 27, 1, 48-57.
Miller, G.A., Galanter, E., and Pribram, K.H. Plans and the structure of behaviour. New York: Holt, 1960.
Popper, K.R. Objective knowledge: An evolutionary approach. Oxford: Clarendon Press, 1972.
Schmitt, M.C., and Newby, T.J. Metacognition: Relevance to instructional design. Journal of Instructional Development, 1985, 9, 4, 29-33.
Taylor, J.C. A dynamic model of memory for research on human information processing. Instructional Science, 1983, 12, 367-374.
Taylor, J.C. Designing instruction to generate expert cognitive skill performance: Empirical evidence. In G. Evans (Ed.), Learning and Teaching Cognitive Skills. Melbourne: ACER, 1991.
Taylor, J.C., and Evans, G. The architecture of human information processing: Empirical evidence. Instructional Science, 1985, 13, 347-359.
Taylor, J.C., and Thomas, N. A knowledge engineering approach to accounting education. Accounting Education, 1994 (in press).
Tennyson, R.D. An educational learning theory for instructional design. Educational Technology, 1992, 32, 1, 36-41.
Tulving, E. Episodic and semantic memory. In E. Tulving and M. Donaldson (Eds.), Organization of memory, New York: Academic Press, 1972, 382-403.
West, C.K., Farmer. J.A., and Wolf, P.M. Instructional Design: Implications from Cognitive Science. Boston: Allyn and Bacon, 1991.
Wittrock, M.C., and Baker, E.L. (Eds.) Testing and Cognition. Englewood Cliffs, NJ: Prentice Hall, 1991.
Young, M.F. Instructional design for situated learning. Educational Technology Research and Development, 1993, 41, 1, 43-58.
_xyz