The Nature and Role of Complexity in Simulation Performance

Abstract

The objective of this study is to understand the nature and role of complexity in simulation performance. In order to do this, a system dynamics model of the product development pipeline was developed, and an online game based on that model was created. An experiment was run where subject made a series of decisions on one or two variables for many periods, with different levels of information available to them (attribute complexity). Hypotheses were proposed based on the literature. The results confirmed the hypotheses relating degree of difficulty to distance from optimality. Two factors adversely affected the subjects' performance: dealing with more complex information on performance (more than one attribute) and having to make more than one simultaneous decision, i.e. more than one decision per period. The latter condition was more detrimental to performance than the former.