ABSTRACT
This chapter aims to furnish an application, illustrating the implementation of spatial statistical/econometric modelling, that specifically examines spatial dependence concealed in the geographic distribution of urban population density, using 1986 census tract data for Toronto and Ottawa-Hull, and both 1980 and 1990 census tract data for Syracuse. Its results should help illuminate the role spatial dependence plays in geographically organizing urban population, and promote the dissemination of spatial statistical/econometric modelling procedures. The geographic organization of urban population within cities has been a topic of interest in urban studies for many decades, and has been investigated for a wide range of city sizes, types, cultural settings, and points in time. Modelling the geographic distribution of population density in urban areas has been undertaken by many regional scientists during the past several decades. A high level of positive spatial autocorrelation was detected, which accounted for much of the systematic variation uncovered in the geographic distribution of population density in Toronto.
