How Do Criminologists Interpret Statistical Explanation of Crime? A Review of Quantitative Modeling in Published Studies

How well a statistical model predicts crime (generally summarized in a measure of the amount of variance explained, or R2) can be an important measure of the comprehensiveness of a theory in explaining a crime problem or of the extent to which a theoretical perspective provides only a partial explanation for the variability in crime. Thus, variance explained provides one method for assessing the state of the criminological theory in criminology. This chapter discusses variance explained and its importance for evaluating multivariate modeling. It shows how the authors have collected data for the original study, and added to it for the chapter. There is very little consideration of the implications that R2 has for the validity of estimates observed in a study. In the end, the chapter considers the implications of these findings for advancing empirical studies of criminological theory.