System Dynamics Applications / 1 / 1.11

1.11 GROUP MODEL BUILDING

By their very nature, complex systemic problems impact on individuals, groups and organisations in quite diverse ways. The reverse also applies: people individually and collectively behave in ways that serve to compound the complexity. This is what Checkland and Scholes (1999) describe as the complexity of human affairs.

It is this complexity of human affairs that demands we pay careful attention to eliciting what is known about a problem situation and using appropriate tools to discover, or find out about, the problem situation. Further, as we set out to find out more about a problem situation our actions frequently change the problem itself (Mason and Mitroff, 1981: 21-26; McLucas, 2003: 16-42; Vennix, 1996: 3-4).

To enhance likelihood of successful design and implementation of strategies to address complex problems it is essential to have intimate engagement of the client group, those who own the problem and the team setting out to analyse the problem (Andersen and Richardson, 1997; Checkland, 1985; Flood, 1999; Lane, 1990; 1992; 1993a; 1993b; Lane and Oliva, 1998; Linard and McLucas, 1999; McLucas and Linard, 2000; McLucas, 1999; 2001; 2003; Morecroft and Sterman, 1992; 1994; Vennix, 1996). Vennix (1996: 3) explains that the primary goal is not to build the model of the system, but rather to get a group engaged in building a system dynamics model of a problem to see to what extent this process might be helpful to increase problem understanding and to devise courses of action to which team members will feel committed.

This was recognised by Kurt Lewin (1946) in his pioneering of action research in the period immediately following World War II in the USA. Action research, action learning, participative management (Weisbord, 1987), organisational learning (Argyris and Schön, 1978; Senge 1990) and more recently group model building (Andersen and Richardson, 1997; Vennix, 1996) all recognise the need for intimate involvement of stakeholders in finding out about the problem, setting out to solve it and designing and implementing remedial strategies. To this end, working together in groups to build system dynamics models which explore systemic behaviour and informing understanding and fostering learning can be highly effective (McLucas, 2001: 2003).

Modelling should never be an activity conducted behind closed doors or by a technician in a dust coat seated at a computer. Group model building involves action research and action learning activities with clients and a team of modellers in an environment of teamwork, cooperation, and trust. Further, it is rarely possible to analyse or develop strategies for the management of complex systems taking the knowledge and skills of practitioners from a single discipline.

Group model building is built on two main pillars. The first involves team members with skills drawn from a range of disciplines. Inter-disciplinary skills and knowledge are drawn from fields as diverse as cognitive psychology, systems thinking, system dynamics modelling, and systems engineering. The inter-disciplinary team may need to be augmented variously by members having domain-specific knowledge—for example:

• operations research knowledge specific to queuing may be needed for a project examining supply chain problems for the military, a large retailer or international shipping agency;
• reliability engineering expertise may be needed to assist in examining and designing remedial strategies for fleet management or the initial distribution of spares to ensure support to the roll out of a world-wide satellite communications system (McLucas, 1999);
• software development expertise may be needed to assist in analysing and diagnosing problems being experienced in the management of software-intensive projects;
• human resource management expertise may be needed to assist in analysing problems for service-delivery firms or large organisations working to develop improved strategies for manpower management;
• organic chemistry, marine biology and oceanography expertise may be needed to provide sufficient input to a modelling effort which analyses the relationship between nutrient run-off from coastal farmlands and the explosion in populations of marine organisms; and
• hydrology, water resources, and land management expertise may be needed to assist in analysis of problems facing government agencies working to develop strategies for alleviation of dry-land salinity.

The augmentation by specialist knowledge and skill will apply to analysis of problems in virtually any complex problem area. It will also provide the basis for estimating parametric values to be used in models when hard data is unavailable (Linstone and Turoff, 1975; Ford and Sterman, 1998).

The second pillar involves close and continual engagement with the client group in a journey of discovery through action research and action learning.

Multi-disciplinary teams will be engaged in virtually all stages. These include developing the initial top-down formulation of the problem, knowledge elicitation, developing dynamic hypotheses, building of conceptual models, building modules, verification and validation of models, and investigation of remedial strategies and their implementation.