Group model building: adding more science to the craft

Abstract

This article discusses the issue of making group model building interventions more of a science than an art by outlining a number of requirements of a research program. Important elements that are discussed are the various goals of group model building interventions and the components and scripts of an intervention. Then the problem of theory development is discussed, together with a number of hypotheses which the authors suggest need more investigation. The article also discusses issues related to the selection of an appropriate research design, as well as a number of thorny measurement problems. © 1997 by John Wiley & Sons, Ltd. Syst. Dyn. Rev. 13, 187‐201, 1997 (No. of Figures: 0 No. of Tables: 0 No. of Refs: 36) The articles in this special issue have presented a number of standard procedures of group model-building approaches, as well as results from empirical studies aiming to identify the success of modeling with client groups. From these articles it becomes clear that group model building is still more art than science. Research on the effects of group model building is scarce; it focuses on a wide variety of outcomes and variables, and research designs differ quite considerably. Instead of a solid research program creating replicable 1 and cumulative results, we seem to have series of presumptions and hunches being repeated in a descriptive literature with little empirical evidence, certainly lacking any sense of competing propositions or refutability of the claims being made (most often by the practitioners who are using the system-intervention approach). The norm for research seems to be to posit an intuitively grounded hunch about what will work with a group and then to design a facilitated conference process around that hunch. If the interventions are successful (in the sense that paying clients like them and are willing to fund them being repeated), then the hunch is substantiated and the best intuitive practice continues. It seems that legends about what is working grow up around these interventions in what can only be described as superstitious behavior. An example may help here. Those of us who work in the field of system dynamics have long held the belief that one of the great benefits of our modeling craft is that