Heuristic diagnostics for the presence of pure error in computer simulation models

Investigates the effect that the absence of pure error has on the statistical analysis of A. Nozari et al. (1987) and identifies conditions under which the presence of the pure error component in the response is preserved. The authors then give a brief overview of the Schruben-Margolin correlation-induction strategy and the statistical analysis given by Nozari et al. Recommendations are made for ensuring the presence of pure error in the response. An example is given that illustrates how lack-of-fit variation in the response can be misinterpreted as pure error.<
>