Abstract
Despite the fact that evaluation of a simulation model is an important process in systems analysis, the issue of model evaluation has often been avoided because of difficulty. In this study, the concept and approaches concerning model evaluation were reviewed and as a result, following recommendations were proposed. 1) Evaluation of a model should be divided into two processes of verification and validation, and conducted from the following three positions; rationalism, empiricism and positivism. 2) Validation of a model should be attempted by objective methods using data taken from real world systems. 3) Simple regression method and the use of Theil's inequality coefficient are useful as objective methods for validation of a model. 4) In validation process, usefulness of a model should be proven rather than complete validity of the model. 5) The conditions of desirable model are to be applied within wide-range, to be used in various purposes and to be tested with data other than those used in modelling.
