STRUCTURE, DATA, AND COMPELLING CONCLUSIONS - NOTES FROM THE FIELD

Some system dynamics models are more effective than others in changing the thinking and actions of their audiences. In my experience, the mo dels that prove most compelling to clients generally have two things i n common: a potent stock and now structure and a rich fabric of numeri cal data for calibrating that structure. Stock and flow structures foc us attention on the intrinsic momentum of a situation and allow one to track movements of people and things in a clear and systematic way. N umerical data not only help to build a client's confidence in a model, but also can materially affect the final structure and key parameter values of a model. Three examples are presented that demonstrate the s trong inferences one may draw when stock and flow structures are combi ned with sufficient numerical data. System dynamics models should be b uilt on a foundation of straightforward core structures and the full r ange of available evidence. (C) 1997 John Wiley & Sons, Ltd.