ABSTRACT
Capture-recapture methods were originally developed in the ecological setting with the aim of assessing the unknown size of an elusive animal population; however, given their wide applicability, they have been gradually extended to epidemiology and public health. $ is called the ratio plot, and it has been developed as a diagnostic device for the binomial by Bohning et al. The ratio plots for the benchmark data examples seem to suggest, in both cases, the presence of unobserved population heterogeneity. The ratio plot is constant with y for densities belonging to the family of power series distributions; this result is worth discussing for the negative binomial distribution. The ratio regression approach can be widely applied, clearly also to ecological data. The approach can be extended in several ways. An interesting extension is that validation information can be easily incorporated into the ratio regression modeling.
