Abstract for: Dealing with Large Models in Qualitative Group Model Building Projects

During a Group Model Building (GMB) project, model complexity tends to increase over time (variables, causal connections, feedback loops). This poses challenges: visual model complexity making the model as a whole unmanageable for workshop participants; maintaining model ownership by participants when altering visual representations to overcome the previous challenge; saturation may not be reachable (due to constraints of time and money). We first investigated causes of model growth and then applied strategies to overcome the above-mentioned challenges: 1. Small groups working on segments of the model in a world-café manner 2. Integrating results of small groups into one model with the whole group 3. Compensating for the lack of saturation through additional validation and freely articulating policy ideas (without looking at a model) Reasons for model growth included: large system boundary due to project scope; participation of elderly poor asked for a high level of detail to stay close to their lived realities; work between workshops: integration of residual facilitation cards, synthesizing models, validation. Whereas strategy 1 worked well, strategy 2 did not and should rather be done after workshops, if at all. Strategy 3 generated only few new ideas. Facilitators of small groups should familiarize themselves with the model segments used to a high degree. One could try to aggregate/simplify the model more, but would then have to plan more time for doing so in preparation of the workshop. One could also try installing an additional role of a “system boundary watchdog” who would argue against model expansion outside of the agreed system boundary (should not be the facilitator).