In justifying his research on criminal careers, Greenberg focuses on methodological issues. He indicates that the probability interpretation of chi square tests depends on the sample being a simple random sample from a larger population. His empirical tests assume a mixed Poisson model which posits that an offender's arrest or conviction rate has a constant value except for random yearly fluctuations. Greenberg points out that the criminal careers debate has been mainly nontheoretical and that arguments for the superiority of one model over another to explain criminal careers have been primarily empirical. In the absence of a link to theory, it is thus difficult to regard the debate as an important one for criminology except for the challenges it poses to modelers and estimators. Research should be conducted with greater attention to procedures that do not bias the research in favor of one model over the other. An appendix contains mathematical equations associated with Greenberg's Poisson process. 14 references