TRAINER NOTES: COGNITIVE MAPPING TECHNIQUES
These notes provide additional background for trainers on cognitive mapping techniques and their uses. They can also be given to students as additional reading if the trainer feels that this would be useful.
Why are cognitive maps important ?
The terms 'cognition' is used in a variety of different ways in the literature (e.g. Rummelhart and Ortony, 1977; Bartlett, 1932; Schank and Abelson, 1977). Here, 'cognition' is used to refer to the mental models, or belief systems, that people use to interpret, frame, simplify, and make sense of otherwise complex problems. These mental models are referred to, variously, as cognitive maps (Tolman, 1948), scripts (Schank and Abelson, 1977), schema (Bartlett, 1932), and frames of reference (Minsky, 1975). They are built from past experiences and comprise internally represented concepts and relationships among concepts that an individual can then use to interpret new events. This is important because decision-makers have a limited capacity for processing information so that, when dealing with complex problems like innovation, they could rarely process all the information that would be relevant. Their mental models help decision-makers to select information and to decide what actions are appropriate (Weick, 1979). Thus individuals' cognitions may shape organisational decisions, although the extent to which this will occur will depend on the socio-political context and on their ability to influence decisions in their organisation.
What role do cognitive maps play in innovation and implementation ?
The process of technological innovation involves important choices about how to design technological solutions to the particular problems that an organisation faces. This involves decisions about both technical (hardware, software, physical layouts, etc.) and organisational (i.e. structures, procedures, attitudes) systems and practices in the firm (Clark and Staunton, 1989; Scarbrough and Corbett, 1992). There is growing evidence that the success or failure of technological innovation depends crucially on cognitions of key people in adopting organisations because these cognitions shape choices about the design of technological solutions (Weick, 1990; Swan & Clark, 1992). Decisions about whether or not to adopt a particular technology may be based more on the subjective beliefs of particular dominant individuals in the firm, than on any kind of collective 'wisdom' or acceptance of the chosen solution (Isenberg, 1988). This can be especially problematic in the adoption of integrating information technologies since these cut across traditional areas of expertise in the firm and require the commitment of different functional groups for their implementation. For these different stakeholders are likely to have very different perceptions and assumptions about the same technology and about its relevance for their particular problems. Moreover, the influence exerted by different stakeholders is not necessarily correlated with the possession of relevant forms of knowledge.
The technology referred to in this role play, MRP2, is an integrating information technology and case studies have revealed many instances of failure or partial failure with this innovation (e.g. Waterlow & Monniot, 1986). Part of the problem seems to arise when individuals from different functional areas have very different beliefs and expectations about the technology and its implications. For example, some individuals may have knowledge that is extremely relevant to choices about whether or not to adopt an MRP2 system, yet are not able to influence the decision process. Others, for a variety of reasons, may not see MRP2 as a particularly useful solution to their problems and yet are expected to be involved in implementing the technology (see Swan and Clark, 1992). Thus difficulties can occur very early in the innovation process when firms are making decisions about whether or not to adopt the technology and what its implications are likely to be. These problems are often the result of the decision-making process and not just the technology per se. At a very early stage it might be important to find a way of involving individuals with a variety of different experiences, knowledge and beliefs, and accessing this tacit knowledge so that the decision-making process is better informed. Cognitive mapping techniques, discussed next, are a set of tools that have been developed to understand and facilitate this process.
How can we identify a person's cognitive map ?
Cognitive mapping techniques aim to provide a tool for revealing peoples' subjective beliefs in a meaningful way so that they can be examined not only by the individual for whom the map is constructed, but also by other individuals and groups (Eden, 1992). The resultant cognitive map will not represent an entire belief system (this would be impossible) but hopes to portray those beliefs that are held to be most significant by the stakeholders concerned. In this way, valuable knowledge can be entered into the decision-making process that may otherwise remain hidden. Individual decision-makers might be encouraged to reflect on their own, perhaps rather narrow, understanding of a particular problem. This is part of the purpose of using the cognitive mapping exercise in this session.
Another potential use of cognitive mapping techniques is to allow decision-makers to look at maps that have been constructed for other stakeholders so that they can begin to understand and appreciate alternative perspectives on the problem. This insight, however crude, into the way that others are thinking about the problem can encourage negotiation and help to reduce conflict (Eden, 1989). An advantage of cognitive mapping techniques (over, say, simply asking someone what they think) is that they allow knowledge to be externalised in some sort of visuospatial layout that is then open for critical reflection. In this way subjective knowledge can be to some extent 'objectified' and therefore discussed in a less threatening way than direct questioning. In this way, if used as the decision-making process unfolds, cognitive mapping techniques may help the process to be managed.
The cognitive mapping exercise that was developed for this session was not used to manage the decision-making process directly, but instead aims to help participants understand and analyse the effect of stakeholders' tacit knowledge, beliefs, and assumptions on decisions about technological innovation. The cognitive mapping technique that you used here was developed specifically for this purpose and is, therefore, necessarily quite crude. However, recent developments in cognitive mapping have led to the availability of a variety of more sophisticated, and in many cases more complex, mapping techniques. The purpose of the rest of this section of the notes is to provide a brief overview of these techniques and to refer the reader to detailed sources of information, should they wish to use or adapt these techniques in their own organisation.
What other applications are there for cognitive maps ?
Mapping techniques use different methods to elicit from an individual key concepts and relationships among concepts and to construct a map for that individual. In this way, differences among individuals can be revealed and, hopefully, understood. This was the purpose of the mapping technique developed for this exercise. Other techniques go further than this and combine maps from different individuals into a composite map that represents the beliefs of a group (e.g. Bougon, Weick & Binkhorst, 1977; Eden, 1992). This is important because beliefs that are shared by a decision-making group have more potential to influence the strategic direction of a firm (Schwenk, 1988).
Cognitive mapping techniques for organisational analysis include: simple content analysis of text (Birnhaum-More and Weiss, 1990); the use of repertory grid techniques (Reger, 1990); the systematic coding of cause and effect relationships (Bougon, Weick 7 Binkhorst, 1977; Axelrod, 1976); special interviewing techniques (Bougon, 1983); computer software analyses of interview data (Eden, 1989); and argument mapping (Fletcher and Huff, 1990). many of these techniques and their applications are reviewed in an edited book by Huff (1990) and in journal articles (see (e.g. Eden, 1992; Swan & Newell, 1994; Swan, 1995). These mapping techniques vary significantly in complexity and in the depth of information they elicit. Some (e.g. argument mapping and computer analysis) are quite complex a time consuming but reveal a lot of detailed information whilst others (e.g. content analysis or the type of exercise you used here) can be carried out fairly quickly but reveal less detailed information. In practical terms, then, there will be a trade-off between the complexity and usability of a mapping technique.
All cognitive mapping techniques reveal concepts that people hold to be important but they vary in terms of the nature of the relationships among concepts that they identify. Some only look at simple categories whilst others aim to reveal deeper underlying arguments (see Figure 1). At the surface level, techniques such as content analysis identify key concepts by looking at how frequently particular words are used in written or verbal statements. Key concepts are taken to be those that are stated most often and key themes can be explored by looking at how words are juxtaposed and at how word usage changes over time. Standard word-processing software can be used to help with this kind of analysis. The problem with these techniques is that they assume that people will volunteer important concepts in their normal use of language. However, in organisations where individuals have different degrees of power and influence, and different vested interests, this may not always be the case.
At a deeper level, techniques such as the repertory grid technique can be used to identify both the content and the structure of an individual's personal frame of reference (Kelly, 1955). Repertory grid techniques work by first identifying concepts and then clustering concepts together in order to reveal underlying dimensions (Reger, 1990). In this way important concepts are identified (e.g. someone might believe: 'MRP2 takes time to implement' and 'MRP2 provides a lot of information') and superordinate categories are revealed (e.g. 'MRP2 is a complex technology'). Elicitation of concepts using repertory grids is fairly straightforward but the analysis of the grids is quite complex and requires factor analysis (with statistical software support). Reger (1990) used a repertory grid to compare the beliefs of different stakeholders and found that there was a relatively low level of agreement among decision-makers as to what constituted an important strategic dimension. However, the technique was not really developed for this purpose (see Kelly, 1955).
At a deeper level still are causal mapping techniques. These aim to identify the key elements of a person's beliefs about a particular problem and to describe the cause and effect relationships among these elements. The exercise used here was modelled on existing causal mapping techniques. The mapping of cause-effect relationships has attracted particular attention, mainly from strategy researchers but also from practitioners. This is because peoples' beliefs about the causes and effects of events are assumed to create expectations about the likelihood of future outcomes and therefore predict strategy formulation and decision-making (e.g. Lyles and Schwenk, 1992). It is worth noting here that this assumption is questioned by some authors (e.g. Ginsberg, 1989). It would be simplistic, for example, to assume a direct relationship between cognitions of individuals and organisational decisions because there will be a number of other contextual and political features that mediate such a relationship. The role play should have illustrated some of these. For example, not everyone would have had equal power or equal opportunity to influence the decision. The beliefs of some stakeholders would carry less weight than others given their status and influence within the organization.
What are the specific techniques employed here ?
The technique used to develop the exercise for this session is very loosely based around a causal mapping technique, SODA (Strategic Options Diagnostic Analysis). This technique aims to reconstruct, from statements elicited during guided interviews with stakeholders, a visual representation of their beliefs about causes and consequences of particular events (for example, the causes and consequences of adopting MRP2). The eventual map produced might look something like the one produced in this exercise but with many more concepts and many more relationships among concepts. A key difference between SODA and the technique used in this session is that with SODA the stakeholders themselves provide the concepts, ensuring that all concepts are personally relevant. This means that individuals can produce entirely different maps for the same problem domain. Another feature of SODA is that the concepts identified are bipolar (e.g. 'carry on with existing systems' as opposed to 'adopting something new') Thus each concept acquires meaning by virtue of a contrast with its opposite pole. Relationships are indicated with arrows connecting those concepts that are causally related and a sign attached to the arrow to suggest whether the relationship is direct (i.e. A causes and increase in B) or inverse (A causes B to decrease). Like repertory grids, concepts can be clustered statistically. SODA is supported by computer software (COPE) that assists with the construction, modification, and various statistical analyses of the maps. The software also has the capacity to aggregate maps so that maps can be produced for groups of stakeholders. This group map would highlight beliefs that are shared by individuals as well as showing those that are idiosynchratic and is used to facilitate teams in making decisions.
The SODA causal mapping technique is used primarily in consultant-client situations where consultants are facilitating group decision-making. The consultant constructs the map following an interview, or a series of interviews, with the client during which the consultant listens to the language used by the client and their assertions and statements about why certain events have occurred. Thus the validity of the technique depends to a large extent on the relationship between the consultant and client to the extent that "the model may be a correct or incorrect representation depending upon the listening skills of the consultant" (Eden, 1988, p. 5).
The consultant also has an important role in helping the clients to examine and reflect on the map and in doing so gain a clearer understanding of the problem. The tool is meant as a device for reflecting on problems, not for solving problems directly. This is typical of cognitive mapping techniques and distinguishes cognitive mapping from other decision support tools. It would be through this process of reflection that individuals or groups could become aware of alternative ways of understanding the problem and then this might then facilitate problem solving. Cognitive mapping techniques do no aim to prescribe solutions. Rather, they aim to encourage the decision-maker(s) to reveal and reflect on their own perception of the problem and to appreciate the problem from others' perspectives so that this may then facilitate the process of decision-making. The mapping exercise used here also had this purpose
A rather different type of causal mapping technique is the grid technique developed by Bougon, Weick and Binkhorst (1977). With this technique individuals are presented with a grid in which concepts are listed in the rows and the columns (see Figure 2). The concepts are usually drawn from initial interviews with respondents (though see Swan and Newell, 1994). For each cell in the grid the individual is asked to consider the nature of the relationship between the row variable and the column variable and, if it is causal, to indicate this in the cell (e.g. does A cause B, B cause A, and is the relationship direct or inverse). The analysis is relatively straightforward and involves the calculation of an indegree score and an outdegree score for each variable in the grid. The outdegree score is the number of paths leading from a variable to other variables and is, therefore, a measure of how important that variable is seen to be in causing change in other variables. The indegree score is the number of paths leading to a variable in the grid from other variables and is, therefore, a measure of how much that variable is influenced by other variables. Further, grids of different individuals can be summated to allow calculation of indegrees and outdegrees for a group. The group map would then represent shared beliefs about the variables that are primarily causal or those that are primarily effects. This kind of mapping technique has been used to identify what individuals or groups believe to be the main causes and effects, for example, of innovation (Swan & Newell, 1994) or quality of performance (Bougon et al, 1977). One disadvantage of the grid technique is that a group map is an average of the individual maps and therefore only represents shared beliefs. This means that individual differences in cognitions which, as seen, might be very important for the decision-process, tend to be obscured in the combined map.
This section of the study notes has suggested that it is important to understand the effect of stakeholders cognitions (tacit knowledge, beliefs, assumptions) on decisions about technological innovation. This understanding needs, however, to be set in the wider social, political and organisational context surrounding decisions about innovation design. Different individuals and groups may dominate the decision process at different times. The context for the role play decision was a decision about adoption of a particular technology. The cognitive mapping exercise allowed participants to critically reflect upon and understand the dynamics of the process that led to the decision to adopt or reject MRP2 as an innovative solution. It also aimed to give participants practical experience of the kinds of cognitive mapping techniques that could be used to facilitate the decision process. The overview of cognitive mapping techniques presented here provides additional information on the variety of techniques available, their uses and their limitations.
Axelrod, R. (1976). Structure of Decision: the Cognitive Maps of Political Elites. Princeton NJ: Princeton University Press.
Bartlett, F.C. (1932). Remembering. Cambridge: Cambridge University Press.
Bougon, M., Weick, K. & Binkhorst, D. (1977). Cognition in organizations: an analysis of the Utrecht Jazz Orchestra. Administrative Science Quarterly, 22, 606-639.
Bougon, M. (1983). Uncovering cognitive maps: the Self-Q technique. In Morgan, G. (ed), Beyond Method: Strategies for Social Research. Beverly Hills, CA: Sage.
Clark, P.A. & Staunton, N. (1989). Innovation in Technology and Organization. London: Routledge.
Eden, C. (1992). On the nature of cognitive maps. Journal of Management Studies, 29, 261-265.
Eden, C. (1989). Using cognitive mapping for strategic options development and analysis (SODA). In J. Rosenhead (Ed.), Rational Analysis for a Problematic World. Chichester: Wiley.
Fletcher, K.E. & Huff, A. (1990). Argument mapping. In A.Huff (Ed.), Mapping Strategic Thought. Chichester: Wiley.
Ginsberg, A. (1989). Construing the business portfolio: a cognitive model of diversification. Journal of Management Studies, 26, 417-438.
Huff, A. (1990). Mapping Strategic Thought. Chichester: Wiley.
Kelly, G.A. (1955). The Psychology of Personal Constructs. New York: Norton.
Lyles, M. & Schwenk, C.R. (1992). Top management strategy and organizational knowledge structures. Journal of Management Studies, 29, 155-174.
Minsky, M. (1975). A framework for representing knowledge. In P.H. Whinston (Ed.), The Psychology of Computer Vision. New York: McGraw-Hill.
Reger, R.K. (1990). Managerial thought, structures and competitive positioning. In A. Huff (Ed.), Mapping Strategic Thought. Chichester, Wiley
Rummelhart, D.E. & Ortony, A. (1977). The representation of knowledge in memory. In R.C. Anderson, R.J. Spiro & W.E. Montague (Eds), Schooling and the Acquisition of Knowledge. Hillsdale NJ: Erlbaum.
Scarbrough, H. & Corbett, J.M. (1992). Technology and Organization: Power, Meaning & Design. London: Routledge.
Schank, R.C. & Abelson, R.P. (1977). Scripts, Plans, Goals, and Understanding. Hillsdale NJ: Erlbaum.
Schwenk, C. (1988). The cognitive perspective on strategic decision-making. Journal of Management Studies, 25, 41-55.
Swan, J.A. & Clark, P.A. (1992). Organisational decision-making in the diffusion and appropriation of technological innovation: cognitive and political dimensions. European Work and Organizational Psychologist, 2, 103-127.
Tolman, E. (1948). Cognitive maps in rats and men. Psychological Review, 55, 189-208.
Waterlow, G. & Monniot, J. (1986). A study of the state of the art in computer-aided production management. Report for ACME, SERC.
Weick, K.E. The Social Psychology of Organizing (1979). Reading MA:Addison & Wesley
Weick, K.E. (1990). Technology as an equivoque: sensemaking in new technologies. In P. Goodman & L. Sproull, Technology and organizations. San Fransisco: Josey Bass
EXERCISE SUPPORT MATERIALS