Can we learn from simplified simulation models? An experimental study on user learning

Abstract

ABSTRACT Simple models are considered useful for decision making, especially when decisions are made by a group of stakeholders. This paper describes an experimental study that investigates whether the level of model detail affects users’ learning. Our subjects, undergraduate students, were asked to solve a resource utilisation task for an ambulance service problem. They worked in groups under three different conditions, based on the type of simulation model used (specifically a simple, adequate or no model at all), to analyse the problem and reach conclusions. A before and after questionnaire and a group presentation capture the participants’ individual and group attitudes towards the solution. Our results suggest that differences in learning from using the two different models were not significant, while simple model users demonstrated a better understanding of the problem. The outcomes and implications of our findings are discussed, alongside the limitations and future work.