Integrating modelling and data analysis in teaching discrete event simulation

The growing popularity of stochastic discrete event simulation in areas such as telecommunication, combined with much marketing hype about ease of use, has coaxed some practitioners into a misguided belief that choosing prefabricated components from libraries and configuring them into a model by pointing and clicking is all that is needed. While neglect of statistical aspects of simulation has already led to some highly problematic published results, this erroneous assumption must also be guarded against in university teaching. This paper therefore argues for the importance of teaching those issues that critically affect the analysis and credibility of a simulation's results alongside those methods and tools targeted at the needs of model design and construction.