Demand Model Estimation and Validation

The overall objective of the Urban Travel Demand Forecasting Project is to provide transportation engineers and planners with the information necessary to select and use policy-oriented disaggregate behavioral travel demand models, and to assess the applicability and limits of specific alternative models. This volume is devoted to the investigations of demand, forming the core of this project.
The research plan underlying this demand research was to:
Collect data on a sample of individual commuters in the SanFrancisco Bay Area before the initiation of Bay Area Rapid Transit (BART) service;
Predict BART patronage from demand models fitted to the pre-BART data;

Compare the predictions with actual BART patronage, using a second survey taken after BART was in service.
Attention was concentrated on work mode-choice. A number of parallel questions in demand analysis were addressed:
 What variables influence demand?
How does the method of measurement of variables affect demand model estimates?