Models based on the application of disaggregate behavioral theories and concepts to travel demand modeling are outlined, and problems associated with their application are discussed. The basic hypothesis of these models is stated, and the inference of values of time from mode choice and route choice models is seen to greatly depend on the accuracy and adequacy of the models. A number of methodological and conceptual problems are posed in the achievement of these objectives. A basic problem that demands attention is the determination of how the phyotheses on which the mathematical techniques are based relate to the hypotheses of choice behavior. An empirical analysis and evaluation of the logit, probit, and discriminant analysis techniques and their underlying mathematical assumptions have revealed the problem of determining a basis for comparing models from different statistical technique. The importance is stressed of establishing statistical validity and confidence in the coefficients of the model variables and of ensuring that the interpretation of the coefficients is not made on the basis of extrapolating the results beyond the range of data. Comments are made on the behavioral interpretation of the coefficients, the time difference coefficients, alternative methods of dealing with user characteristics, and impurities relating to other differences among the modes or routes. The problems of the specification of the models for different treatments of trip segments are discussed, and operational problems associated with theoretical model structures for applying logit analysis to a multiple-choice situation are reviewed.

Media Info

  • Media Type: Digital/other
  • Features: Figures; References;
  • Pagination: pp 12-18
  • Monograph Title: Value of travel time
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 00141939
  • Record Type: Publication
  • ISBN: 0309025532
  • Files: TRIS, TRB
  • Created Date: Apr 13 1977 12:00AM