Abstract
In recent years considerable efforts have been made to develop models of urban and regional systems but very little practical advice has been produced to help planners use these models in the most appropriate fashion. The paper describes a statistical methodology for using models in planning exercises which takes into account their imperfections and the need for an explicit link with policy objectives. It also provides a new approach to investigating the empirical behaviour of models of applied significance.

This publication has 14 references indexed in Scilit: