Lessons Identified from Data Collection for Model Validation

Computer simulation models will ideally be developed within an environment where all the necessary input data is readily available and all the relevant stakeholders are supportive and co-operative. In practice, many models are developed for the purpose of evaluating organisational or process change. As a result, system performance data may often be limited or potentially biased by key parties associated with the results of the simulation model. Ensuring valid input data is therefore a key aspect of model validation and an important component in a successful simulation study. This paper addresses some examples of data collection problems that have been encountered by the author and presents a number of lessons identified