Comparing different sampling schemes for approximating the integrals involved in the efficient design of stated choice experiments

Abstract

The semi-Bayesian approach for constructing efficient stated choice designs requires the evaluation of the design selection criterion value over numerous draws taken from the prior parameter distribution assumed when generating the design. The semi-Bayesian D-criterion value of a design is then calculated as the average value of the D-errors over all the draws taken. The traditional way to take draws from a distribution is to use the Pseudo-Monte Carlo approach. However, other sampling approaches are available as well. Examples are Quasi-Monte Carlo approaches using Halton sequences, Faure sequences, modified Latin hypercube sampling and extensible shifted lattice points, a Gauss–Hermite quadrature approach and a method using spherical-radial transformations. Not much is known in general about which sampling scheme is most efficient for calculating semi-Bayesian D-errors when constructing efficient stated choice designs. In this study, we compare the performance of these approaches under various scenarios and identify the most efficient sampling scheme for each situation. The method based on a spherical-radial transformation is shown to outperform the other methods when small numbers of draws are used.