ABSTRACT
Data analysis is a key element of computer simulation. Factors such as arrival rates and processing times affect the performance of a process, and they are typically random in nature. Field data can be used on a proposed redesign of an existing process, but it is likely that the problem of not having enough data to draw meaningful conclusions would be encountered. Probability distribution functions include these extreme values in the tails. Model validation is necessary before the model can be used for analysis and prediction. A valid model also can be used for buy-in with the process stakeholders; that is, if the stakeholders are convinced that the model does represent the actual process, then they are more likely to accept recommendations that are drawn from testing design alternatives using the computer simulation. The analyst can choose the probability distribution based on how well each of the tested distributions fits the empirical distribution drawn from the sample data.
