ANALYZING LOCATIONAL ATTAINMENTS - CONSTRUCTING INDIVIDUAL-LEVEL REGRESSION-MODELS USING AGGREGATE DATA

The authors address a major difficulty associated with research on spatial processes. Although theory frequently demands analytic models that combine variables from individual and locational (or aggregate) levels, appropriate data sets are rarely available. The authors demonstrate that, in many circumstances, it is nevertheless possible to construct individual-level multivariate models by using aggregate census files and microdata samples, and they describe a method for doing so. They illustrate this method with an analysis of suburban location in the metropolitan area around New York City, focusing on two features of suburban communities-their homeownership level and median household income.