This paper is an examination of functional form and functional misspecification in regression-based spatial interaction models. Box–Cox transformations are used to model US state-to-state labor migration. Attention is given to the production-constrained gravity model, the competing and intervening destinations (CID) model, and an extension of the CID model which uses dummy variables and an iterative transformation procedure. Statistical evidence supports the argument that the use of highly restrictive log-linear specifications may be inappropriate and problematic. Implications for spatial interaction modeling using general nonrestrictive functional forms are discussed.
No takes yet. Share an insight, caveat, or question.
Fik et al. (1998) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: