EXTERNAL STRUCTURE OF VARIABLES IN INDIVIDUAL CHOICE MODELS OF TRAVEL DEMAND

Despite the recent developments in the modelling of individual choice, especially in transport applications, where economic theory has been successfully used in the derivation of an operationally tractable choice model, there has been relatively little consideration of the specification of the explanatory variables beyond a distinction between the generic and alternative-specific classification and the level of measurement (i.e. reported-perceived or engineering measures). Gaudry and Wills (1977) have recently examined the functional form of an entire system of travel demand models, and Louviere (1978) has identified strong evidence for the manner in which variables might be entered into a function (i.e. as additive, multiplicative etc). In the present paper, we discuss a related aspect, the specification of the "external" structure of a variable' i.e. whether it should be specified as a ratio, a transformed ratio, a difference etc. The economic interpretation of alternative external structures is summarised; and then no matter which sturcture is chosen, we show that the variables may enter the choice model using the more general Box-Cox transformation instead of the "restrictive" linear funcitonal form (although for difference configurations a restricted range of translation parameters must be used). The alternative configurations are empirically examined in a simple binary-mode choice model using the Box-Cox transformation. /Author/

  • Corporate Authors:

    Macquarie University, Australia

    School of Economic and Financial Studies, Balaclava Road
    North Ryde, New South Wales,   Australia 
  • Authors:
    • Hensher, D A
    • JOHNSON, L W
  • Publication Date: 1978-5

Media Info

  • Features: Figures; References; Tables;
  • Pagination: 19 p.

Subject/Index Terms

Filing Info

  • Accession Number: 00177316
  • Record Type: Publication
  • Report/Paper Numbers: Res. Paper No. 159
  • Files: TRIS
  • Created Date: Jul 19 1978 12:00AM