Stated Preference methods and travel demand forecasting: An examination of the scale factor problem

Stated Preference choice models may have incorrect scale properties, because their coefficients incorporate the influence of factors which do not affect actual behavior, and this paper examines the consequences of this scale factor problem for demand forecasting. Empirical evidence is presented which suggests that the scale factor problem is not serious and the results are generalised to consider the consequences of the scale factor problem in situations other than those prevailing in the empirical example. The paper also examines the predictive performance of three procedures for forecasting changes to an existing demand in the light of the scale factor problem. This indicates that reliance should not be placed on the incremental logit model and that a simple procedure which amends a standard logic model forecast in the light of its performance in explaining the observed initial demand is generally preferred to unadjusted logit model forecasts.