Abstract
The paper provides a framework for judging various spatial choice and interaction models on the basis of methodological, theoretical, logical and practical criteria. After a brief discussion of several common models (such as gravity models, entropy models, logit models, probit models and multiproportional Poisson models), the significance of these models is assessed by confronting them with the above-mentioned criteria, so that a qualitative evaluation can be made.
The logit model which turns out to be a rather satisfactory spatial interaction model, is dealt with in more detail. Especially the problems of aggregation and disaggregation are analysed, while also a new contribution to logit models is provided by constructing a general probability density function for the population in case of more than two choice alternatives.
Get full access to this article
View all access options for this article.
