sensitivity analysis

A formalized procedure to identify the impact of changes in various model components on model output. Sensitivity analysis is an integral part of simulation experimentation and may influence model formulations. It is commonly used to examine model behaviour. The general procedure is to define a model output variable that represents an important aspect of model behaviour. The values of various inputs of the model are then varied and the resultant change in the output variable is monitored. Large changes in the output variable imply that the particular input varied is important in controlling model behaviour. Within this general definition, sensitivity analysis has been applied to a variety of model inputs including state variables, environmental variables and initial conditions. (Source: YOUNG)