Validating the Assumptions of Sequential Bifurcation in Factor Screening

Abstract Sequential bifurcation (or SB) is an efficient and effective factor-screening method; i.e., SB quickly identifies the important factors (inputs) in experiments with simulation models that have very many factors—provided the SB assumptions are valid. The specific SB assumptions are: (i) a second-order polynomial is an adequate approximation (a valid metamodel) of the input/output function of the underlying simulation model; (ii) the directions (signs) of the first-order effects are known (so the first-order polynomial approximation is monotonic); (iii) so-called “heredity” applies; i.e., if a specific input has a “small” first-order effect, then this input has “small” second order effects. Moreover, SB assumes Gaussian simulation outputs if the simulation model is stochastic (random). A generalization of SB called “multiresponse SB” (or MSB) uses the same assumptions, but allows multiple types of simulation responses (outputs). In this article, we develop heuristic practical methods for testing whether these assumptions hold, and we evaluate these methods through Monte Carlo experiments and a case study (namely, a Chinese logistics network).