Scribd Logo
Upload

Welcome to Scribd!

  • 366 views

    Types of Knowledge

    Uploaded by

    ndhoat
    differentiate types of knowledge to be learnt

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd

    Types of Knowledge

    Uploaded by

    ndhoat
    366 views9 pages
    differentiate types of knowledge to be learnt

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    differentiate types of knowledge to be learnt

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Download as pdf
    366 views9 pages

    Types of Knowledge

    Uploaded by

    ndhoat
    differentiate types of knowledge to be learnt

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Download as pdf
    You are on page 1of 9
    EDUCATIONAL PSYCHOLOGIST, 31(2), 105-113, Copyright© 1996, Lawrence Erlbaum Associates, ne Types and Qualities of Knowledge Ton de Jong University of Twente Monica G.M. Ferguson-Hessler Eindhoven University of Technology Research in learning and instruction claims a central role for the concept of knowledge. The knowledge base of a person, it is now generally assumed, is made up of different types of knowledge. The most well-known examples are declarative and procedural knowledge, but, ‘more elaborate distinctions exist Furthermore, the knowledge base is characterized by different ‘Ualities, such as level (deep or surface) of knowledge, generality of knowledge, level of ‘aulomization of knowledge, modality of knowledge, and structure of knowledge. The present article examines the concept of knowledge by presenting a matrix that takes nypes and qualities ‘of knowledge as its dimensions. Ths matrix can be used to classify research on knowledge by linking aspects such as knowledge assessment techniques, expert-beginne differences, instruc- tional measures, and leaming goals tothe cells of the matrix. In literature on learning and instruction, knowledge plays a pivotal role and is attributed a wide varity of properties and, qualities. Among the examples encountered are generic (or general) and domain specific knowledge, concrete and ab- stract knowledge, formal and informal knowledge, declara- tive and proceduralized knowledge, conceptual and proce- dural knowledge, elaborated and compiled knowledge, ‘unstructured and (highly) structured knowledge, tact or inert, knowledge, strategic knowledge, knowledge acquisition Knowledge, situated knowledge, and metaknowledge. For ‘example, in an article by Reif and Allen (1992), atleast eight different knowledge terms are used: main interpretation knowledge, general knowledge, definitional knowledge, an- cillary knowledge, supplementary knowledge, case-specific knowledge, entailed knowledge, and concept knowledge. In an earlier article on the same topic (interpreting scientific concepts), Reif (1987) also used the terms declarative know!- edge, procedural knowledge, formal knowledge, compiled knowledge, special knowledge, general knowledge, proce- dural interpretation knowledge, and coherent knowledge. Ap- parently, researchers need many and fine-tuned terms for describing the knowledge stat of individuals, ‘A numberof studies have signaled this explosion of con- structs and terms and have undertaken to structure the field, approaching the various aspects of knowledge from a general "Requests fr reprini shouldbe sent to Ton de Fong. Facalty of Educ tional Selene and Technology, Univesity of Twem, P.O. Box 217, 7500 [AE Enschede, The Netheslands.E-malljong@edteutwene cognitive perspective (Alexander & Judy, 1988; Alexander, Schaller, & Hare, 1991; Snow, 1989).In this article, we also, attempt to give a systematic description ofthe various aspects ‘of knowledge, but we approach this goal from the perspective of knowledge-in-use. This means that task performance forms the basi forthe identification of relevant aspects of knowl- cedge. In this approach, we have found it effcientto introduce ‘wo dimensions that describe knowledge: type of knowledge and quality of knowledge. ‘We consider these dimensions tobe independent, and we make a systematic distinction between characteristics speci fying the ype of knowledge (e.g, conceptual knowledge) and characteristics specifying the properties or qualities of knowledge, which can in principle be relevant for several types of knowledge (e.g, modality). We think this distinetion is necessary to avoid the introduction of still more types of ‘knowledge that do nothing more than describe properties of generally acepted types of knowledge. An example of such an introduction can be found in Jonassen, Beissner, and Yacci (1993), who introduced “structural knowledge ... the knowl- edge ofhow concepts within a domain ate interrelated” (p.4). The idea of structural knowledge could be described more parsimoniously asa combination of knowledge type or types (eg. conceptual knowledge) and quality (structure). In this article we demonstrate the general importance of the concepts of type and quality of knowledge for theory, research, and practice in te field of leaning and instruction. By using the two concepts type and quality of knowledge as dimensions, a matrix canbe created that can structure general 106 D6 JONG AND FERGUSON-HESSLER topics such as learning goals and expert-novice differences. The studies we review here are restricted in the sense that we focus mainly on one type of criterion task, problem solving, and on domains from the sciences, particularly physics. TYPES OF KNOWLEDGE Frequent attempts have been made to give a systematic de- scription of knowledge. Some attempts have been based on cognitive theories, whereas others have been formulated to serve as a basis for instructional design theories. Still another approach is to characterize knowledge from an epistemologi- cal point of view. This implies that elements ofthe knowledge base are characterized by the function they fulfill in the performance of a target task. Epistemological approaches are task dependent. This means that different classifications (on- tologies) of knowledge types are contrived for different types of tasks (Gott, 1989) and that within one domain, the same elements of subject matter may be characterized by different ‘ontological typologies, depending on the task in which they funetion (de Jong, de Hoog, & Schreiber, 1988; Messick, 1984), Numerous authors introduce distinctions between types of knowledge, often without taking into account these charac- teristics of epistemology. From the perspective of cognitive theories, intrinsic in the description of knowledge i aclassi- fication that has absolute qualities. Alexander and Judy (1988), for instance, in their review article on the interaction ‘of domain specific and strategic knowledge, distinguish three types of domain specific knowledge: declarative (factual in- formation), procedural (compilation of declarative know!- edge into functional units that incorporate domain specific strategies), and conditional (understanding when and where to access certain facts or employ particular procedures). In a later review article by Alexander et al. (1991), this same distinction is used, and reference to the task is only made in the term ask knowledge, understanding ofthe cognitive de- mands of a task. The same type of absolute classification is foundiin instructional design theories, of which Gagné (1985). Merrill (1983, 1987), Reigeluth (1983), and Romiszowski (1981) are renowned authors. Central elements of this type of theory are objectives and criterion tasks, which are typical for domain, However, a variety of types of knowledge is introduced, including concepts, principles, and procedures, without reference to their function in complex task perform ance, Although absolute classifications may serve a general ‘goal, we think a more pragmatic typology of knowledge takes into account the context in which the knowledge has to function. Our own work (e.g., de Jong & Ferguson-Hessler, 1984, 1986, 1988, 1991; Ferguson-Hessler & de Jong, 1987, 1990, 1993)has been inthe field of physics, where task performance inthe form of problem solving and experimental work plays 4 central role. Knowledge of physics is characterized by strong links between elements, a high degree of abstraction, and a hierarchical nature. In other words, expert knowledge ‘of physics is strongly and hierarchically structured, and atthe lower levels of the hierarchy local chunks are tailored to applications. ‘The explicit focus of our work has been on problem solving. On the basis of a detailed task analysis we distin- ‘guished four types of knowledge: situational knowledge, conceptual knowledge, procedural knowledge, and strategic knowledge. In defining these types of knowledge, we make use of an example from mechanics (see Figure 1) to clarify the differences between knowledge types. This problem is a typical example of exercises usually found in first-year uni- versity textbooks on physics. It is the same type of problem used by Chi and Bassok (1989); Chi, Feltovich, and Glaser (1981); Hammer (1994); and Zajchowski and Martin (1993), Situational knowledge is knowledge about situations as they typically appear in a particular domain. Knowledge of problem situations enables the solver to sift relevant features, out of the problem statement (selective perception) and, if necessary, to supplement information in the statement (Braune & Foshay, 1983). It may serve to create a repre- sentation of the problem from which, ifthe organization of knowledge is adequate, additional knowledge (conceptual, procedural) can be invoked. In Figure 1, examples of situ- ational knowledge could be knowing that a rough surface means frictional force, which acts against motion, or know- are other forces working on the blocks than the tly mentioned—for instance, a normal force from ‘A wooden block of mass m is slidin which mak the plane by means of a string, plane. The string runs. hanging freely from the string. between pl acceleration of each of the blocks. angle « with the horizontal. The block is pu fara pulley at the top of the plan ‘at its other end, it supports a second block of mass M, which is. The coefficient of dynamic friction ne and block m is i. Find the direction and size of the ‘rough, inclined plane, FIGURE 1 Example problem fom mechanics Conceptual knowledge is static knowledge about facts, concepts, and principles that apply within a certain domain. Conceptual knowledge functions as additional information that problem solvers add to the problem and that they use 10 perform the solution. In our earlier work we used the term, declarative knowledge. Now we prefer to use the term con- ceptual knowledge (see also Greeno, 1978), and, following vvan Berkum and de Jong (1991), we use the term declarative as a quality characteristic, being the opposite of compiled knowledge. An example of conceptual knowledge in Figure 1 could be knowing that the size of the friction force is the coefficient of friction times the normal force, or knowing that the total force vector acting on a body equals its mass times its acceleration, Procedural knowledge contains actions or manipulations that are valid within a domain. Procedural knowledge helps the problem solver make transitions from one problem state to another. Itcan have a specific, domain-bound (strong) char- acter, or itcan be more general (Weak). In the example presented in Figure 1, procedural knowledge concems knowing how to identify and delimit each of the two interacting mechanical systems, how to choose a coordinate system, and how toresolve the forces acting on the blocks in the chosen system, Strategic knowledge helps students organize their prob- Jem-solving process by directing which stages they should go ‘through to reach a solution. A strategy can be seen asa general plan of action in which the sequence of solution activities is Iaid down Posner & McLeod, 1982). Elements of knowledge belonging to the first three types are specific, applicable to certain types of problems in a domain, whereas the lat type, strategic knowledge, is applicable to a wider variety of types of problems within a domain (Bransford, Sherwood, Vye, & Rieser, 1986; Polya, 1957). In Figure 1, strategic knowl- ‘edge concerns knowing how to organize and interpret the information given, structure it in a diagram, define the mechanical system(s) to be used in the analysis, distin- ‘guish internal and external forces, list all external forces acting on cach of the system(s), and translate the result- ing force diagrams into equations that can be solved for the acceleration, ‘The previous analysis illustrates the importance of episte- mological domain analysis for instruction. A proper classifi- ‘cation of relevant types of knowledge may be used to organize and present subject matter and to direct the process of learn- ing. For a full description of the knowledge base, however, we need to be able to distinguish not only various types of knowledge, but also the quality of the components of the knowledge base. QUALITIES OF KNOWLEDGE ‘A large number of concepts are used to describe qualities of knowledge: Generic, abstract, informal, elaborated, and struc- tured are but a few examples. Some qualities refer to relations between knowledge types, whereas others refer to types as "TYPES AND QUALITIES OF KNOWLEDGE 107 such, Some qualities are more suited for specific types of, knowledge (e.g., compiled for knowledge of proce- dures), whereas others are used in a general way (e.g., depth). Although we discuss a number of qualities sepa- rately, it should be noted that in many cases there seems to bbe some overlap. Again, we illustrate various qualities of knowledge with elements from the mechanics problem in Figure 1. Level of Knowledge: Deep Versus Surface Level of nowledge is term that is used loosely in literature and in educational practice. Mostly, a rough distinction is made between swface or superficial knowledge and deep knowledge, with the connotation that deep is good and surface is poor. Knowledge is called deep when itis firmly anchored ina person's knowledge base and when external information has been translated to basic concepts, principles, or proce dues from the domain in question. Such knowledge is differ- ent from the concrete appeerance of the external information from which it tems. Decp-level knowledge is associated with comprehension and abstraction, with critical judgment and evaluation, and the like (see, ¢g., Marton & Saljo, 1976). This knowledge has been thoroughly processed, structured, and stored in memory ina way that makes itusefl for application and task perform- ance (Glaser, 1991). Even so, the general structure may well have idiosyncratic features. Snow (1989) described the de- sired end states of learning in terms of “articulated, deep “understanding of a domain, including the ability toreason and explain in causal terms, and to adopt multiple viewpoints about a problem or phenomenon” (p. 9). Surface-level know! ‘edge is associated with reproduction and rote learning, trial and error, anda lack of critical judgment (Glaser, 1991). This knowledge is stored in memory more of less as a copy of external information. Larkin (1983), when discussing experts’ knowledge rep resentations, made a distinction between a basic repre sentation (made up of concrete objects from a problem state- ‘ment, a technical representation that includes concepts from the domain), and a computational representation (made up of equations). In a later study, based on the well-known study by Chi et al. (1981), Chi and Bassok (1989) described the problem representations of students of mechanics in the fol- lowing terms: ‘The poor problem solver has a basic repre- sentation consisting of explicit entities from the problem description (the surface model), whereas the successful prob- lem solver has a physics representation that includes, in addition, generated physics entities not explicitly described (the deep model), The physics representation, as mentioned by Chi and Bassok, is similar to Larkin's technical repre- sentation and reflects the application of a deep level of knowledge Looking at the solution of our mechanics problem from Figure 1, we identify surface elements such as the rough 108 0c JONG AND FERGUSON-HESSLER inclined plane, wooden block, string (elements of situational knowledge), or searching for a formula containing aecelera- tion (as part of strategical knowledge). The corresponding deep clements would be something like an object sliding up an inclined surface pulled by a string and slowed by a frietion force (situational knowledge), or knowing how to draw a force diagram for one ofthe blocks and translate this diagram into an equation (elements ofa strategy). Strategic knowledge offers well-known examples of differences between superfi- cial and deep clements of knowledge. Superficial strategies are, for instance, an algebraic search for a formula containing the unknown entity, or filling in given quantities in a formula. Deep clements of strategic knowledge are, for instance, ana- |yzing and interpreting the information given, structuring this information by means of a diagram, and explicitly defining the physical system that is being analyzed. Structure of Knowledge One of the first researchers to stress the role of the structure of knowledge in memory was Larkin (Larkin, McDermott, Simon, & Simon, 1980). She concluded from her studies on problem solving in experts and novices thatthe large amount of knowledge stored in the memory of an expert is made possible not by a general superiority of memory, but by the chunking of information into large, meaningful units—a type ‘of organization that most novices lack. These results are con- sistent with the early results found by Chase and Simon (1973) ‘onthe perception and memory of chess masters, in which groups of stimuli were perceived and remembered as units. ‘The importance of structuring knowledge has been stressed by Reif (1984; Reif & Heller, 1982) and others (Camacho & Good, 1989; Elio & Scharf, 1990; Prawat, 1980), Reif stated that a hierarchic knowledge structure isthe type most suited for retention of knowledge, for quick and efficient search processes, and for fitting in new elements of knowledge without restructuring knowledge already present. ‘This type of structure, which contains abstract and general concepts at the higher levels, is typical of expert knowledge (see also Boshuizen & Schmidt, 1992). A more extended notion of the organization of knowledge that includes not only ‘conceptual knowledge but also other types of knowledge can bbe found in the concept of schema, Knowledge of a domain includes several schemata, tailored to typical task perform- tance in the domain and organized in a systematic, often hierarchic way (Rumelhart, 1980). A schema contains the different types of knowledge required for task performance, In the case of problem solving, we can speak of problem schemata containing situational, conceptual, and procedural knowledge (de Jong & Ferguson-Hessler, 1986) that corre- spond to a particular problem type, The role of an efficient knowledge structure was demonstrated in an experiment where we investigated the knowledge structure of novice, first-year students of physics. The results showed that stu- dents who were good problem solvers (i.c., showed expertise at their own level) had a knowledge structure that was more similar toa set of problem schemata than students who were poor problem solvers (de Jong & Ferguson-Hessler, 1986). In fa theoretical analysis of the subject matter of a first-year course on electricity and magnetism (Ferguson-Hessler & de Jong, 1987), we demonstrated how a set of problem schemata can be used as a basis for a hierarchic structure, forming the Iowest level and being subsumed into the hierarchy as more knowledge is acquired. Itis evident from this discussion that depth and structure of knowledge are not independent. Only the introduction of deep elements makes possible the generalizations and abstrac- tions that are required for the construction of (problem) schemata and the building of a hierarchical structure. How- ever, itis possible to build a structure on superficial charac- teristics, and such a schema would also be remembered and used in applications. It might contain incorrect elements and relations and would thus lead to faults in the application, Such 1 structure could be described as a noncanonical schema, tacks the functionality of schemata belonging to a well- structured knowledge base (Taconis & Ferguson-Hessler, 1993) ‘The relation between depth and structure is found in stra- tegic knowledge as well. Superficial elements can be chained into astrategy we named “kick-and-rush” Find a formula fil in, calculate, ready (de Jong & Ferguson-Hessler, 1984). A ‘well-structured strategy contains a logical series of actions, such as analyzing information, constructing a problem repre sentation, selecting tools for the solution, and planning the various stops to be carried out to reach the solution. Together, these actions form a strategy, but they cannot be isolated from that strategy and carried out independently. In the mechanics problem, we can distinguish elements of conceptual knowledge that ae either loosely connected (e.g, force equals mass times acceleration, friction equals 1 times normal force, normal force equals normal component of ‘weight)or structured ina logical way e-g., the net force acting, ‘onasystem equalsits mass times the acceleration ofthe center ‘of mass, thus in case of linear motion, net force normal to the direction of motion equals zero). ‘Automated (Compiled) Versus Nonautomated Knowledge ‘Task performance of a beginner can be a conscious, step-by- step process of choice and execution based on fairly general ‘methods, For an expert this changes into a continuous, fluid, ‘and automatic process based on strong domain or situation specific methods, a deep representation of the given task, and well-structured knowledge of principles and procedures. Such a knowledge base is described as being compiled—that is, tailored for acertain type of application (Anderson, 1983). Hereby attention is freed forthe continuous parallel checking, Of the task execution, which is characteristic of expert per- formance. The total effect of automation is fast and reliable task performance. The distinetion between automated and nonautomated knowledge is strongly related to the distinction between tacit ‘and explicit knowledge. Wagner (1991), following the Ox- ford English Dictionary, defined racitnovledge as “practical ‘know-how that is usually not openly expressed or stated” (p. 173). According to Wagner, it “typically is acquired through informal learning, ether from one's own experience or from that of a mentor or colleague” (pp. 173-174). Gelman and Greeno (1989) denoted this (ype of knowledge as implicit knowledge. Broadbent and colleagues were among the first to emphasize the existence of tacit knowledge (Bey & Broadbent, 1984, 1988; Hayes & Broadbent, 1988). Tacit knowledge is seen by these authors as knowledge that is not ‘open to being verbalized. In a series of experiments, Berry and Broadbent (1984) had subjects interact with a simulation (e.8.,controlling a sugar factory), gave them aspecific assign- ment (eg, t0 reach a certain level of sugar production), and {questioned them afterward. Results from this study showed that the ability to answer questions on a written posttest was rot correlated with the ability to control the factory success- fully. Apparently, people possess knowledge that enables them to perform a task, but it i tacit, implicit, or not easily ‘expressed. On the other hand, explicit knowledge is not by

    You might also like