ABSTRACT
An alternative method of modelling the urban system is to develop a model based on linear regression and econometric techniques. So far only linear models with one independent variable have been considered, but it would obviously be far better to include several independent variables using multiple regression analysis. One important aspect of the calibration of a multiple regression equation model is in determining the significant independent variables, that is the variables that are important in explaining the zonal variation in the dependent variables. Regression analysis is a highly developed statistical technique and there is a set of basic assumptions related to its use. Unlike the gravity model where there are no statistical assumptions, a linear urban model based on regression analysis is faced with a stringent set of assumptions. With the type of socioeconomic data used in urban modelling studies, there are often problems in satisfying these assumptions.
