The magnitude of individual-level variations in demand coefficients: A Xenia, Ohio case example

The research reported in this paper develops individual value and choice expressions for a sample of travellers in Xenia, Ohio, U.S.A. A brief description of the methodology for constructing and forecasting with individual-level demand equations is given, relying principally on multiple linear regression and multiple analysis of variance. The research tests the hypothesis that coefficients of individual value functions are systematically associated with experiential, situational and personal characteristics of the individuals. We also obtain data on the distributions of the individual value coefficients and their interrelationships. Finally, the research applies these specifications to the retrospective prediction of transit patronage in a situation of rapidly changing transit system characteristics from 1974 to 1977.