Exploratory data analysis is a procedure used to gain qualitative insight rather than to test hypotheses. Thus, it differs from current approaches in the social sciences. The exploratory method examines the first few samples to make a qualitative assessment of the data. Then it plots relationships between variables to arrive at various functions used in later modeling. Before constructing a regression model, simple bivariate correlations are checked to identify highly collinear variables and to exclude unimportant variables. The output of each analytical stage is the input for the next stage. The exploratory data analysis approach is applied to crime statistics in the New Jersey suburbs of Philadelphia. Results show that property crime is best explained by the attractions of wealth and commercial establishments these localities provide to outside offenders. Accessible communities are more attractive than those lacking urban access. It is suggested that if this statistical association is verified, the results have implications for crime control activities in metropolitan regions. (Author abstract modified)