Simulation As a Necessary Step in the Design of Stated Preference Experiments



Simulation As a Necessary Step in the Design of Stated Preference Experiments

Authors

TUDELA A M, Universidad de Concepcion, Chile

Description

Stated Preference is a technique widely spread between transport practitioners and researchers to carry out market studies and supporting decision making. It originated in the marketing discipline, where it was known as Conjoint Analysis, then was transfe

Abstract

Stated Preference is a technique widely spread between transport practitioners and researchers to carry out market studies and supporting decision making. It originated in the marketing discipline, where it was known as Conjoint Analysis, then was transferred to the transport field in the 80s. Well-documented reviews about the development of this technique can be found in Bates (1988), Hensher (1994) and Ort~izar (2000).

One of the steps that is suggested when designing a Stated Preference experiment is the simulation of responses (Fowkes and Wardman, 1988). This would allow the designer to improve the quality of the design. This step has been adopted as a code of practice by many people, but with results that are seldom reported in the literature.

This paper argues that simulation is actually an unnecessary step in the design process due to the ambiguity of its results, than can be adjusted by the designer to obtain any value he/she might want to. This subjectiveness reduces the validity of the procedure.

Two approaches were adopted to analyse the previous hypothesis. The first one is theoretical, based on the principles used to generate the simulated responses and estimate the coefficients, whereas the second is a practical one. The latter uses an experimental design, showing that data can be manipulated by the researcher to recover any value that might be desirable.

Basic aspects related to the simulation process are reported in the next section, whilst theoretical aspects of the analysis are shown in section 3. The practical approach is reported in section 4, whereas the main conclusions are pointed out in section 5.

Publisher

Association for European Transport