Process and outcomes of modeling: an attempt at formulating a conceptual framework

Abstract

In the field’s early publications, Forrester was already clear on the goals of the system dynamics approach. The purpose of constructing a system dynamics model of any real world issue of concern was to improve system functioning. The vehicle for doing so was an increased understanding of relations between structure and behavior or a change in mental models, leading stakeholders to implement conclusions of the modeling effort. In recent years much attention has been given to involving stakeholders directly in modeling, leading to various forms of group model building. Involvement in modeling was expected to increase change in mental models and thereby foster implementation of conclusions. However, only a few years ago leading authors in the field have called group model building more an art than a science, and pointed to the lack of evidence for its effectiveness. Authors in related fields have made similar comments with regard to their respective methodologies: a lack of an integrative conceptual framework linking modeling practice to modeling outcomes, makes 1. an accumulation of research results difficult and 2. leaves implicit what the crucial elements of a modeling intervention are. Thus the effectiveness of methodologies is difficult to assess and relate to relevant theories. In this article such an integrative conceptual framework is proposed, drawing on theories of persuasion (mental model change) and the influence of beliefs and evaluations on actions. Well-established theories in social psychology are used to formulate important variables and relations in the model, enabling us to draw on standard operationalizations and relate to a body of research on social interactions. The framework is used to evaluate five group model building cases in a PhD research project, of which the main results are reported here. The results indicate that group model building effectiveness can be understood in terms of the framework, leading to the identification of further research questions.