DISCRETE CHOICE MODELS: ISSUES, DATA AND APPLICATION

There is a well established body of literature which emphasises the uniqueness of individuals (or households) in the development of quantitative tools to explain and predict choice behaviour. The approach is dominated by the assumption that choices are made amongst mutually exclusive (and collectively exhaustive) alternatives, and that the heterogeneity of the sampled population must be recognised at the outset and built into model specification. The resulting models have been characterised as individual discrete-choice models, typified by the family of random utility models (multinomial logit, nested logit, generalised extreme value, binary probit, multinomial probit, dogit) estimated with disaggregate data sets. This paper highlights some key elements of this approach, and illustrates their present use in the area of shopping centre (destination) choice. We distinguish use of the technique in spatial-aspatial settings as well as the nature of data (revealed-stated preference/choice). Examples are drawn from the extant literature. (Author/TRRL)

  • Corporate Authors:

    Australian Institute of Urban Studies

    Acton, A.C.T.,   Australia 
  • Authors:
    • Hensher, D G
  • Publication Date: 1984

Media Info

  • Features: Figures; References;
  • Pagination: p. 147-175

Subject/Index Terms

Filing Info

  • Accession Number: 00392330
  • Record Type: Publication
  • Source Agency: ARRB
  • Files: ITRD, TRIS, ATRI
  • Created Date: Mar 29 1985 12:00AM