Opportunity to improve: the first large scale empirical analysis about the replication of system dynamics studies

Organizational and management sciences have continuously fail to replicate existing studies. Moreover, replication the hallmark of scientific inquiry is not investigated adequately. The few existing studies examine pertinent questions about scientific integrity since errors in publications are more commonplace rather than single instances. Since computational simulation modeling is increasingly employed in studies in organizational and management sciences, we investigate the replicability of simulation studies, especially of the system dynamics type. Based on a sample of 105 studies in scholarly management and organizational journals we conclude that fundamental limitations exist in replicating existing models. Only in 20% of the cases, is an identical replication possible. Often the necessary details to replicate the simulation are not available or the available details are incomplete. Most often, it seems, this incompleteness is a surprise to the original authors as well. We provide suggestions to improve the degree of replicability to ensure full replicability is a standard for system dynamics simulations models in organizational and management sciences.