Printer Friendly

Developing consensus: reflections on a model-supported decision process.

What roles do model-based decision support systems play in the achievement of consensus in management teams?


McGregor has listed a number of characteristics observed in the operation of effective teams[1]; the list includes:

* The members listen to one another; every idea is given a hearing

* The group is comfortable with disagreement, showing no signs of having to suppress conflict resulting in "open warfare", or the need for resolution through a vote.

* Decisions are reached by a kind of consensus in which it is clear that everybody is in general agreement and willing to go along

Clearly not all teams charged with making a collective decision are fortunate enough to possess all these characteristics throughout their deliberations. This article is concerned with how consensus might be achieved in management teams, specifically by investigating the role that model-based decision support systems (DSS) can play in this process. The models involved are detailed computer-based models that are used to simulate or "run forward" a system -- whole industry sectors in these cases -- for typically periods of three years up to 30 years in order to gain strategic insights of likely industry behaviour and to evaluate strategic options.

The author has been involved in a number of consulting assignments for groups in which the conceptualization, development and use of complex competitive models of such businesses played a central role. In each case the objective was ostensibly to evaluate a single or limited number of identified strategic options -- which they successfully did -- but in fact the process was commended by the clients most strongly for the way it had lead to a level of consensus in the group that had not been expected.

Webber and O'Connor[2, p. 20] have remarked that, while some studies have found that DSSs impacted positively, in the majority there is no discernible effect. They further commented, "It is difficult to trace the reasons for these mixed results." These conclusions suggest that little may have changed in most applications since Martin and Winch[3], whose survey similarly showed little evidence for the successful application of model-based support systems in senior-level strategy-related decision making They further concluded that the potential for failure of such systems would remain unless behaviourally acceptable processes were created. The objective of this article is to reflect on the activities involved in the consulting project described and to relate them to theories and perspectives of modelling and of consensus building in an attempt to trace some of the reasons for their evident success.

The article opens by recognizing that, while there is evidence that teams working collectively can make better decisions, the team members inevitably have individual ideas and views that have to be aligned as a prelude to the evaluation of decision options. It continues by reviewing different perspectives of the role of models to support group deliberations. A particular approach to group decision support is described in which the whole process of conceptualization, development and use of a system dynamics-type model plays a key role, and one example case-study is recounted in which the model's evolution and use was able to contribute significantly to the development of a shared view. The best exploitation of an investment option and the resulting benefits were evaluated in such a way that the team was able to reach a consensus-based decision.

The article concludes by reflecting on the process described in the case in the context of conditions specified for effective group decision making and the particular attributes of the modelling approach used. Parallels are drawn with other group decision aids that aim to raise consensus, and in this way the particular features of this approach that contributed most strongly to consensus development are identified. These various projects, of which the described case is typical, were all undertaken in a strict commercial environment, and the hypotheses could not be checked by trials or control experiments, only by reference to underpinning theories and other published trials. It must be borne in mind though that the majority of controlled experiments on such aspects of computer-supported decision making have been undertaken with student or other volunteer groups working in simulated rather than live decision-making situations[2].

Attributes in the process described do emerge that accord strongly with results from those experiments and with the views of other authors in pointing to why the achievement of consensus in these studies did come about.

"Shared views" and the role of computer models

A key feature of senior-level decision making is that it is rarely undertaken by one manager alone. Even when one individual carries ultimate responsibility, he or she will rely on colleagues to provide inputs to the formulation of options and the evaluation process, and will rely on the team to ensure successful implementation. Indeed as organizations have become more complex, progressively fewer decisions are made by single individuals[4], and it has been long argued that with the right internal processes groups can produce very high quality decisions, often better than those that could be expected of members of the same group acting as individual decision makers[5]. Offering suggestions for the "right" processes, Cooke and Slack[6] again have included such factors as:

* a high level of participation by all group members;

* equal status in the treatment of ideas, whatever the source;

* a consensus approach in the taking of the decision, where genuine attempts are made to secure the agreement of all group members to the decision that is finally arrived at.

In such group decision-making circumstances each member of the team will start with a "view" of the situation. A view is the knowledge, information, and data the manager possesses about the company, its competitors, the business/industry, the macro-economic environment and expectations of the future. Each executive's view is based on and developed from his or her individual background and experience, an individual knowledge base, personal access to information and data, the unique interaction with others, understanding of the processes and dynamics of the business, and the individual's selectivity in interpretation. Each executive's view is, therefore, different from all the others, imperfect and incomplete.

Each view is highly complex and wide-ranging and usually extremely difficult to articulate. However, a shared view of the key driving forces in the business is essential to the management team if they are to move forward to formulate and implement the most successful strategies. The need for a shared view is accentuated in a modelling process by the fact that there will eventually be a single basic model, and for it to have maximum impact and utility it must command the confidence of the greatest number of key executives -- preferably all.

The use of models to support decision making is common-place, and, particularly since the emergence of the personal computer, these models have progressively been used more and more by decision makers themselves. This is in contrast with earlier times when the interaction with models was undertaken remotely by specialist analysts. Great emphasis has been placed on developing systems with easy-to-use interfaces to produce models that may easily be accessed by managers (e.g. Bonczek et al.[7]) and to enable the output from the models to be assimilated easily. A number of specific developments have emphasized the potential role of supporting groups rather than individuals.

The emerging concept of the group decision support system (GDSS), whereby systems and facilities are established specifically to support a group or team in analysis and decision-making sessions, may also include model bases among their facilities[8,9]. In such systems, the group may assemble in a purpose-built room in which each member sits at a terminal to interface with databases, model-bases, display options, and with one another through GDSS applications software. Alternatively an LDN (local decision network), teleconferencing, or even a remote decision network may be utilized. Such systems permit the sort of climate advocated by Cooke and Slack, particularly as the system may display inputs from individuals publicly while keeping their source anonymous, thereby ensuring equal treatment of all ideas. (However, this approach may lead to increased conflict as people perhaps become more assertive or even ruder than they would be face-to-face.)

Advanced screen graphics have been identified as a powerful tool for communicating information to others and industry reports (e.g. Lembersky and Chi[10] point to significant industry savings attributed to graphics. Interactive visual decision making exploits such computer-based graphic displays to show interactively the impact of different decisions, through graphical summaries of data and/or iconic displays of systems or processes. Many such systems are developments from queue simulation-type approaches, but are based on a visual model that is used as an integral part of the decision-making process[11-13]. The need for managers to take on trust output from simulation modelling, rather than being involved centrally in the development of scenarios and seeing how the simulated results unfurl was identified by Turban[14] as a major potential cause of credibility gaps and dissatisfaction with simulation modelling. Highly visually stimulating interfaces and interactive experimentation may also be expected to facilitate the speed of exploration and the opportunity for managers to review the evaluations in small groups rather than just individually[11,15; see also 16,17] in terms of influence diagrams and belief nets; and Eden[18], for computer-assisted group development of cognitive maps). Certainly everyone familiar with computer arcade games can appreciate the speed and versatility of such systems -- for example SimCity -- and indeed the potential of such technology in management learning through computer-supported cases as recognized in the term "edutainment"[19,20].

In the above examples of the potential contribution of models in group decision making, the main, if not exclusive, attention has been on the use of a previously developed model to facilitate evaluation of options. From this the team can develop an agreed strategy based on the simulated outcomes, but there will have to have been universal or at least majority trust in the model and its underlying assumptions. The modelling and analytical approach known as system dynamics, which is used in the situations in this article, does, however, offer the management team a much greater opportunity to participate in model development not just model use (see, for example, Morecroft et al.[21].

Modelling as the catalyst in consensus building: a case study

The modelling approach -- system dynamics

System dynamics is an approach to understanding and predicting the dynamic behaviour of complex business, social and other systems through a process of "mapping" the structure of the system and then simulating its behaviour over time with an explicit computer-based quantitative model[22-24]. The approach itself was originally conceived[25] to be transparent and participative, and the modelling software developed for the approach has been intentionally easy to understand by the non-specialist[26-28]. Fundamental to the approach is the assumption that there are underlying structures in real-world systems which predetermine their behaviour, and that these structures can be captured, and system behaviour can be simulated with explicit mathematical models. The approach utilizes simple diagramming techniques to map out system structure and assumptions, providing the team, admittedly generally, with the assistance of a specialist, to appreciate the complex interactions in the systems and identify creative decision options[29].

The model development process has been characterized most recently by the technique's originator as a six-step process as summarized in Figure 1[30]. The process is intended to capture the complexity, non-linearity and feedback loop structures that are inherent in social and physical systems in terms of the flow and accumulation of entities in the system -- money, people, resources, information and other system state variables like motivation, awareness, and propensity to act or buy and so on. Appropriate software products include DYNAMO, STELLA, ITHINK, POWERSIM, which, alongside other products with similar basis and/or syntax, facilitate formulation of these structures into a dynamic simulation model. They provide easy access to model output, and a number also permit animations which can further enhance appreciation of the dynamic behaviour of the system.

Situation background

The case study that is described briefly here as an illustration relates to the petrochemical industry, in which the modelling was undertaken to support a client's strategic decision making The client company concerned is a major producer of speciality and semi-commodity chemical and petrochemical products, and was at the time evaluating a major investment that could lead to a significant reduction in manufacturing costs for one of its product businesses. The increased margin gain that would be generated was originally seen as the basis for the justification of the investment expenditure. However, an alternative view arose, namely that the cost saving be used to reduce product price, thereby enabling the company to "buy" market share -- it would achieve lower margin gain but, taken over a larger volume, even better returns could potentially be attained. The project was ostensibly initiated by the client in order to support their simple evaluation of the particular investment option but, as it emerged, there was not uniform acceptance of the purpose to which improvements in process economics would be put, nor the cost savings that would result. As will be seen, the whole process of model conceptualization, construction and scenario evaluations enabled the management team in the client company to agree on not only the value of the investment but also how the resulting manufacturing cost gain should be used.

The core simulation model was constructed using the system dynamics approach and was written in DYNAMO[23]. The model captured the process parameters, economics and decision making in the industry and its marketplace. The modelling process was "extensive"[31] in the sense that the model evolved during wide-ranging conversations and data gathering from key players in the decision team over a period of time. (Other users of the approach have successfully utilized an "intensive" approach -- a single, focused meeting of all players that might last one or more days; see, for example, Anderson and Richardson[32]). Structural diagrams, as shown in Figures 2 and 3, were used during this process; these proved particularly useful in helping the individual players to clarify structures in their minds, and to articulate them and share their views of the world with their colleagues.

Before examining the details of the modelling and consensus-building process, some general features of this sort of competitive model are described.

The general nature of the class of competitive models

This consulting assignment was typical of a number which have been undertaken by the author, particularly in the chemical and petrochemical industries, centring on the development and use of very large system dynamics models. These models have used subscripted model sectors to represent, on an individual basis, clients and competing producers and the markets they serve. The models inevitably differed in a number of respects depending on such factors as the nature of the product, for example whether it was a pure commodity or produced in a range of grades or formulations, the diversity of the process technologies in the industry, and whether a single or a number of countries or geographic regions were being analysed. Occasionally the models also incorporated more complex upstream/downstream subsystems for interrelated products. However, they also had a number of features in common, and Figure 4 gives an overview of the generic structure of these models.

The models generally comprise four main sectors, which are outlined below.

Competitive producers

This sector consists of an array of variables forming a generic representation capturing the key structural relations and decision processes in the operations of the client and each of the major competitors. These formulations cover a wide range of aspects:

* Nameplate capacity and plant utilization.

* Capacity expansions -- "roundout" and new base capacity -- and shutdowns.

* The investment decisions related to capacity and other forms of strategic developments.

* Impact of production technologies on such aspects as yield, feedstock requirement, grades of product, and operating economics.

* Cost reduction owing to "learning curve effects", general competitive responses, or particular strategies.

* Full representation of the operating economics and associated accounting: fixed costs, variable operating costs, feedstock costs, cash margins, profits, financial performance measures.

* Market targeting; in differentiated products, the directing of output towards various end-uses and markets based on marketing strategies, relative profitabilities, and product capabilities.

End-use demand

This is generally a relatively simple sector inputting forecasts of basic demand for the various grades or formulations of products, but which may include representations of price/demand effects, and product substitution.

Market interface

This is a particularly important sector in models which represent differentiated products and regional markets. The sector brings together the outputs from the various competitors, targeted according to individual capabilities and priorities, and reconciles them with the various end-use demands. The interface determines the relative success of the competitors in each market given their output, price and performance characteristics.

Product pricing

The pricing mechanism generally operates at two levels:

(1) A general price level for the product is set on the basis of the overall potential supply/demand situation.

(2) Prices are also determined for the various grades or markets, given any differential production costs, yield issues and the individual balance of supply (targeted product) and demand.

Figures 2 and 3 show typical diagrams used to represent the structure of particular aspects of these subsystems. They show alternate representations of the manufacturing capacity, Figure 2 being in causal-loop diagram or influence diagram format and Figure 3 in stock/flow diagram format. These are standard formats associated with this approach, though each has its own proponents[30,33-35]. Both formats show here, for example, that productive capacity can be increased either through the addition of new base capacity or by increasing the effective capacity of existing plant - "incremental" or "roundout" capacity. Strategists will focus their capacity expansions on one or other, or both, of these at any time depending on the investment economics and other factors - for example, they are more likely to go for new capacity at times when demand growth projections are strong, plant utilization is near 100 per cent, and prices are bullish. Capacity may similarly be taken out of operation but "moth-balled" to be restarted if the plant economics change, or finally decommissioned as it becomes obsolete. The diagrams also helped in surfacing and sharing such received understanding as that the supply/demand balance influencing product price may also be impacted on by shutdown capacity - i.e. the moth-balled plants that could be restarted - and not just the plants that are currently fully operational.

A range of parameters are specified to enable all these features to be characterized for each producer and/or product. Other mechanisms may also be included to permit the investigation of particular scenarios and options, for example mergers or acquisitions among the competitors, new entrants or withdrawals. The models thus comprise true "bottom-up" representations of the industry sectors with the aggregate industry built up from individual producer and market sectors with the inevitable complex influences and feedbacks affecting the decisions made within it.

Evaluation of the investment option

The complex model was constructed on a bottom-up basis as described above, aggregating individual competing producer sectors to represent the industry as a whole. This model was then used to simulate futures for the industry that accounted for the competitive manoeuvrings of the different producers to evaluate the client's project from both perspectives and under various possible macro-economic futures. The model included full representations of producers' investment decisions, processes, technology enhancement, and pricing as indicated above. The complexity that the model was able to incorporate spanned the areas of responsibility and expert knowledge of a number of the division's senior functional managers. It also utilized a number of generic model structures that the author and colleagues had developed that represented mechanisms and processes typical of such industries. Such generic structures speed up model building generally but, just as importantly, they enhance the acceptance of the model by allowing the incorporation of detailed mechanisms in terms of accounting or operating procedures, technology learning curves and buyer sensitivities without incurring large dollar or time costs[36]. The author has also argued that the incorporation of such established generic structures may be considered as constituting a form of hybrid knowledge-based system[37].

The model itself achieved a high level of confidence across the range of managers concerned - each feeling that his/her knowledge and perspective of the industry had been incorporated, and also appreciating that each of the options could be fully investigated in the ensuing analysis.

The model was ultimately used in a series of experiments to evaluate the impact of the different options in terms of the dynamic future of the industry, the likely competitive reactions of other producers, and the resulting performance of the client company. The model suggested that a price cut by the client of the size envisaged would result in severe retaliatory price cutting by competitors, and insufficient increased volume would be achieved to warrant the loss of margin. However, further sensitivity analysis showed a policy of passing on of the order of 40 per cent of the cost reduction to customers to be optimum, achieving some gain in market share without instigating an aggressive price war. Just as important, in the longer term this policy would generate higher returns enabling further technical improvements to the product, thus maintaining competitive advantage as competitors responded in due course with cost reduction investment of their own. Having moved to a shared view on the optimum way to exploit the potential cost savings from the process enhancement, the investment appraisal could be completed.

Intervention by the modelling approach thereby not only provided a quantitative evaluation of the decision options, but also was arguably of much greater value in aiding the team to come to a consensus view on how the firm's strategy should be pursued in order to exploit to the full the benefit that would accrue. It is probable that, even if the precise "answer" for optimum pricing strategy proved inaccurate, the company would still have gained greater advantage from the investment through implementing the decision on the basis of the shared commitment within the team, than if the project had proceeded against a backdrop of disagreement.

Key attributes of the approach that fosters consensus

Model development and model use

Earlier, an overview of the modelling process was given, and in this section aspects of that process, as applied in this typical case, are related to other research and theories on consensus building and DSS benefits. The process described was basically linear in six stages, but with many cycles back to refine and review as the project progresses. In this section the first two steps are considered together as the model development phase, and the next three taken together as the model use phase; the sixth step of implementation of new policies is not addressed here.

A categorization of media available to support a negotiation process has been offered by Poole et al.[38] within which the process described here falls closest to their "level 2 computer support" classification. This category was originally defined in terms of GDSS[39] which, in addition to a modelling facility, also utilize the computer system for basic GDSS functions such as anonymous entry of ideas and preferences and voting solicitation. The process undertaken here did not utilize those other facilities but did put great emphasis on the model development process, rather than simply offering the group the use of a model to evaluate options. The potential benefits within the phases of development and use are thus highlighted separately here.

The model development phase

A key feature in the modelling process in this case was the drawing-up of structural diagrams that captured the basic structures in the system. Various forms of iterative process, many leading to the development of some form of visual model, have been claimed to assist in group consensus building, including the following:

* The Delphi technique is the classic iterative process, and nominal or focus group techniques are also iterative, but none normally leads to any form of explicit model.

* Interpretative structural modelling (ISM) uses a (usually) computer-supported voting sequence to develop a structural model which comprises a ranked order of priorities that define an issue problem[40,41].

* Computer-aided cognitive mapping has also been used as a group technique in management, for which Eden also uses the term strategic options development and analysis (SODA)[9,42]. This process is seen as a way of retaining the richness and complexity of the real world without making it overwhelmingly debilitating, and the emerging model is also seen as a vehicle or dialectic device to facilitate negotiation[43].

* Other computer-supported techniques also seek to add depth to mapping or brainstorming techniques through iterative group-based thinking, for example the use of hexagonal "idons" (the combination of idea and icon) leading to influence diagrams described by Hodgson[44].

The model development phase here mirrors these approaches in bringing together the thinking of the group. Such collaborative modelling building has been credited with inducing further co-operation[45], and the ability to compare multiple views has been valued by Yusoff and Jenkins[46] as an avenue for discoursing about such issues as norm and value systems that need to be incorporated into the problem definition, and ultimately the DSS. Effectively here the issue at hand changed from being "Is the investment justified?" to "How can the potential benefits from the investment be best exploited?" Eden[42] has commented with regard to cognitive mapping that aggregation of maps into a single group map, which will not "belong" to any individual member of the group, means it becomes a device for "facilitating negotiation, synergy, and creativity"; this aspect was strongly reflected here in terms of the development of the structural diagrams and the final model used to evaluate the different options. Moore[47] also argues that a principle for group working is that effectiveness can be enhanced if a group memory - some form of open record of ideas and concerns - is utilized, and the structural diagrams, and perhaps key model formulations, may also be viewed as such.

In this respect it was particularly valuable that the diagramming and flowcharting conventions used here had two key characteristics:

(1) They were close to reality - managers were able to recognize easily the real-world processes they were trying to explain and/or understand.

(2) They readily translated into the code that formed the computer model, so that managers could have confidence that the model was indeed made up of the structures they had identified, and also so that changes could be made simply and transparently.

The final model was complex, but intelligible; it was a single model, but one that reflected the thinking and views of each team member; it captured and interrelated the softer processes involved in the system, but did not leave technical and specialist managers uncomfortable with unfamiliar and subjective characterizations of their worlds. The development phase thus contributed to consensus building in many ways that are similar to that achieved by other group development and consensus building techniques. However, the team had now also participated in the production of the explicit model that would be used to simulate system behaviour to evaluate objectively the strategic options available.

The model use phase

Quantitative simulation models are used because they allow many scenarios to be evaluated in a short period of time, in a controlled manner, and without having to experiment on the real world. However, the availability of the computer-based model of itself offered further possible benefits. It is considered that computer systems may aid the resolution of conflict because they seem impartial and credible[38], while Nyhart and Samarason[45] suggest that the provision of an external objective model with which the parties can work can prevent rigidity (added emphasis). The accessibility of the model development process would appear to have enhanced these aspects of the model's eventual use. It was also of key importance that reruns of the model with parameter or structural changes could be carried out with minimum fuss, preferably by the managers themselves, or at least by the consultants jointly with them.

Generally the positive experiences from model use in this typical case mirror the conclusions of Poole et al.[38] in their review of the impacts observed in studies of various media in negotiation processes. They conclude that the major beneficial impacts of "level 2 computer support" - that is, access to DSS that include the facility to use a model - are:

* surfacing existing conflicts; clarifying procedures and facilitating work on complex tasks;

* providing a common focus;

* increasing the number of ideas considered and providing models for generating solutions; and

* counteracting negative climate and engendering positive emotional expression.

Dangerfield[27] has observed that there has been a shift of emphasis in the role of such group-based models from "content to process", but in this situation the use of the model in the "hard" sense of providing quantitative evaluations of identified options was of central importance. The need for quantitative model outputs had been a key original objective and they added significantly to the confidence the team held in the indicated course of action. The author has discussed elsewhere[48] the circumstances where such "hard" benefits from model support are of value equal to or greater than the "softer" benefits of system understanding or team learning

The circumstances of this model-supported decision mirror closely those reported in Dickmeyer[49], which is one of the few reported controlled experiments on the effectiveness of decision support systems where the test group comprised practitioners operating in their familiar environment. Dickmeyer's work involved semi-structured budget planning by university administrators and it reported significant positive improvement from the use of the DSS. Key features of those circumstances that also applied here were: a sophisticated process model with some representational aids to assist appreciation of output data; a complex task environment; and high task familiarity among the group.


In the case described above, the decision concerning the benefit of the investment, the basis on which benefit should be calculated, and indeed the direction the company would go in exploiting the advantage, became intimately interwoven with the development and use of the model. Divergent prior positions were resolved through this process through the evaluation of the different strategies by the neutral medium of the model, enabling differences to be expressed and evaluated with minimal conflict. The openness of the approach was certainly a major factor in permitting a high level of participation by all group members, and in offering an unequivocal evaluation of ideas regardless of who proposed them. In terms of what both McGregor and Cooke and Slack identified as an important element of the "right" process for effective group decision making, there was a clear consensus approach in the taking of the decision, in which genuine attempts were made to secure the agreement of all group members to the decision that was finally arrived at.

In a review of the experiences of some US and European companies in using dynamic computer models in developing and analyzing strategic scenarios, Tenaglia and Noonan[50] commended the practice of surfacing and testing the implicit mental models of senior managers as a process even more essential than learning, because it helps build consensus and can lead to increased decisiveness. This requires the active involvement of the managers in model development and use, and these should not be "assigned to expert futurists or financial analysts". However, they also insist that the process of using models and scenarios has to be "well done", and this article has described and analysed a process adopted in consulting practice that has proved successful.

Given the prevailing circumstances, a process involving the development and use of complex structural models that could simulate system behaviour proved very effective, and appears to have done so because it accorded with the high task complexity and high task familiarity of the specialist managers involved. However, this does point to a need to understand better the level of richness that might be optimal for different management group types if using such a modelling process - for example, technical specialists like accountants or operations managers might prefer models more rich in detail than general managers or those from softer science backgrounds, and there may be conflicting pressures with widely disparate groups.

In summary, a modelling approach, as represented in the example case here, can make a significant contribution to consensus building in three phases of the decisionmaking process:

(1) The conceptualization of the model focuses attention and draws out a shared view on the key driving forces that determine the future of the industry/business and the company's relative performance.

(2) The creation of an agreed, valid model and the development of a "most likely" outlook for the business and company based on forecasts derived from the model provide a common basis for assessing impacts of the company's future actions and/or of external events (e.g. in the economy or energy prices).

(3) Evaluation of actions and options available to the company by scenario analysis with the model, including quantification of the impacts, leads to an agreed strategy for the company.

[Figure 1-4 and Illustration Omitted]


[1.] Stuart-Kotze, R., Introduction to Organizational Behauior, Reston Publishing, New York, NY, 1980.

[2.] Webber, R. and O'Connor, M., "The effectiveness of decision support systems: the implications of task complexity and DSS sophistication", Journal of Information Technology, Vol. 9, 1994, pp. 19-28.

[3.] Martin, C.J. and Winch, G.W., "Senior managers and computers", Management Decision, Vol. 22 No. 2, 1984, pp. 12-24.

[4.] Gannon, M.J., Organizational Behaviour: A Managerial and Organizational Perspective, Little Brown, Boston, MA, 1979.

[5.] Kelly, H.H. and Thibaut, J.W., "Group problem solving", in Lingron G. and Aronson, E. (Eds), Handbook of Social Psychology Vol. 4, Addison-Wesley, Reading, MA, 1969.

[6.] Cooke, S. and Slack, N., Making Management Decisions, Prentice-Hall International, Hemel Hempstead, 1984.

[7.] Bonczek, R.H., Holsapple, C.W. and Whinston, A.B., "The evolving roles of models in decision support systems", Decision Sciences, Vol. 11 No. 2, April 1980.

[8.] Gray, P. and Nunamaker, J.F., "Group decision support systems", in Sprague R.H. and Watson, H.J. (Eds), Decision Support Systems: Putting Theory into Practice, Prentice-Hall International, Hemel Hempstead, 1989.

[9.] Eden, C. and Radford, J. (Eds), Tackling Strategic Problems: The Role of Group Decision Support, Sage Publications, London, 1990.

[10.] Lembersky, M.R. and Chi, U.H., "Decision simulators' speed implementation and improved operations", Interfaces, July-August 1984.

[11.] Bell, P.C. et al., "Visual interactive problem solving - a new look at management problems", Business Quarterly, Spring 1984.

[12.] Stefik, M.J. et al., "Integrated access-oriented programming into a multi-paradigm environment", IEEE Software, January 1986.

[13.] Hackwell, G., "Discrete event simulation - a commercial perspective" in Moscardini, A.O. and Fletcher, EJ. (Eds), Applied Simulation and System Dynamics, Emjoc Press, Northallerton, 1991.

[14.] Turban, E., Decision Support Systems and Expert Systems: Managerial Perspectives, Macmillan, Basingstoke, 1988.

[15.] Turban, E. and Carlson, J.G., "Interactive visual decision making" in Sprague, R.H. and Watson, H.J. (Eds), Decision Support Systems: Putting Theory into Practice, Prentice-Hall, Englewood Cliffs, NJ, and Hemel Hempstead, 1989.

[16.] Abramson, B. and Finizza, A., "Using belief networks to forecast oil prices", International Journal of Forecasting, Vol. 7, 1991, pp. 299-315.

[17.] Oliver, R.M. and Smith, J.Q., Influence Diagrams, Belief Nets and Decision Analysis, John Wiley & Sons, Chichester, 1990.

[18.] Eden, C., "Cognitive mapping and problem structuring", System Dynamics Review, special edition, Vol. 1 Nos. 2-3, 1994, pp. 257-76.

[19.] Langley, P.A. and Larsen, E.R., "I can't believe it's a system dynamics model", Microworlds, Proceedings of the 1994 International System Dynamics Conference, System Dynamics Society, Stirling, Scotland, 1994, pp. 73-82. See also "Converging technologies: multimedia and gaming simulation", working paper, The Learning Centre, London Business School, March 1994.

[20.] Business Week, "Special report - the learning revolution", 28 February 1994, pp. 42-9.

[21.] Morecroft, J.D.W., Lane, D.C. and Viita, P.S., "Modelling a biotechnology start-up firm", in Milling, P. and Zahn, E. (Eds), Computer-based Management of Complex Systems, Springer-Verlag, Berlin and New York, NY, 1989.

[22.] Wolstenholme, E.F., System Enquiry: A System Dynamics Approach, John Wiley & Sons, Chichester, 1990.

[23.] Richardson, G.P. and Pugh, A.L., Introduction to System Dynamics Modeling with Dynamo, The MIT Press Cambridge, MA (now available from Productivity Press, Portland, OR), 1981.

[24.] Richardson, G.P., Feedback Thought in Social Science and System Theory, University of Pennsylvania Press, PA, 1991.

[25.] Forrester, J.W., Industrial Dynamics, The MIT Press, Cambridge, MA (now available from Productivity Press, Portland, OR), 1961.

[26.] Pugh, A.L., DYNAMO Users Manual, The MIT Press, Cambridge, MA (now available from Productivity Press, Portland, OR), 1983.

[27.] Dangerfield, B., "The system dynamics modelling process and DYSMAP2", European Journal of Operational Research, Vol. 59 No. 1,1992, pp. 203-9.

[28.] Richmond, B. and Peterson, S., ITHINK Documentation, High Performance Systems, Hanover, NH, 1990.

[29.] Flood, R.L. and Jackson, M.C., Creative Problem Solving: Total Systems Intervention, John Wiley & Sons, Chichester, 1991.

[30.] Forrester, J.W., "Systems thinking, systems modelling", System Dynamics Review, special edition, Vol. 10 Nos. 2-3, 1994.

[31.] Richmond, B., The Strategic Forum: from Vision to Operating Policies and Buck Again, High Performance Systems, Hanover, NH, 1991.

[32.] Anderson, D. and Richardson, G., "Scripts for group model building", Problem-solving Methodologies, Proceedings of the 1994 International System Dynamics Conference, System Dynamics Society, Stirling, Scotland, 1994, pp. 25-34.

[33.] Coyle, R.G., Management System Dynamics, John Wiley & Sons, Chichester and New York, NY, 1981.

[34.] Wolstenholme, E.F., "A systematic process to model creation", in Morecroft, J. and Sterman, J. (Eds), Modelling for Learning Organizations, Productivity Press, Cambridge, MA, 1994.

[35.] Richmond, B., "Systems thinking: critical skills for the 1990s and beyond", System Dynamics Review, Vol. 9 No. 2, 1993, pp. 113-34.

[36.] Graham, A., "Generic models as the basis for computer-based case studies", Proceedings of the 1988 International System Dynamics Conference, System Dynamics Society, La Jolla, CA, also available as Report d-3947, MIT Press, Cambridge, MA, 1988.

[37.] Winch, G.W., "Tomorrow, today: system dynamics or expert systems", in Milling, P. and Zahn, E. (Eds), Computer-based Management of Complex Systems, Springer-Verlag, Berlin and New York, NY, 1989.

[38.] Poole, M.S., Shannon, D.L. and DeSanctis, G., "Communication media and negotiation processes", in Putnam, L.L. and Roloff, M.E. (Eds), Communication and Negotiation, Annual Reviews of Communication Research, Vol. 20, 1992, pp. 46-66.

[39.] DeSanctis, G. and Gallupe, B., "A foundation for the study of group decision support systems", Management Science, Vol. 33, 1987, pp. 589-609.

[40.] Warfield, J.N., Societal Systems: Planning Policy and Complexity, John Wiley & Sons, Chichester, 1976.

[41.] Warfield, J.N., "Consensus methodologies", in Group Planning and Problem-solving Methods in Engineering Management, Olsen, S.A. (Ed.), John Wiley & Sons, Chichester, 1982.

[42.] Eden, C., "Strategic thinking with computers", Long Range Planning, Vol. 23 No. 6, 1990, pp. 35-43.

[43.] Eden, C., "A framework for thinking about group decision support systems (GDSS)", Group Decision and Negotiation, Vol. 1, 1992, pp. 199-218.

[44.] Hodgson, A.M., "Hexagons for system thinking", European Journalfor Operational Research, Vol. 59 No. 1, 1992, pp. 220-30.

[45.] Nyhart, J.D. and Samarason, D.K., "The elements of negotiation management: using computers to help resolve conflict", Negotiation Journal, Vol. 5, 1989, pp. 4362.

[46.] Yusoff, M.B. and Jenkins, J., "An approach to development of decision support systems for a public sector administration", Journal of Information Technology, Vol. 9, 1994, pp. 26-37.

[47.] Moore, C.M., Group Techniques for Idea Building, Applied Social Research Methods Series, Vol. 9, Sage Publications, Newbury Park, CA, and London, 1987.

[48.] Winch, G.W., "Consensus building in the planning process: benefits from a 'herd' modelling approach", System Dynamics Review, Vol. 9 No. 3, 1993, pp. 287-300.

[49.] Dickmeyer, N., "Measuring the effects of a university planning decision aid", Management Science, Vol. 29, 1983, pp. 673-85.

[50.] Tenaglia, M. and Noonan, P., "Scenario-based strategic planning: a process for building top management consensus", Planning Review, Vol. 20 No. 2, 1992, pp. 12-19.

Application questions

(1) How important is consensus in effective decision making?

(2) Are models such as those the author discusses appropriate only for very large organizations?

Graham W. Winch is Professor of Business Analysis at the Plymouth Business School, University of Plymouth, Devon, UK.
COPYRIGHT 1995 Emerald Group Publishing, Ltd.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1995 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Winch, Graham W.
Publication:Management Decision
Date:Nov 1, 1995
Previous Article:Strategy development: past, present and future.
Next Article:Decision processes, monitoring, incentives and large firm performance in the UK.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters