Statistical support for identification using epigenetic traits of the human skeleton

Radiologic comparison is a potentially reliable means of identification in forensic contexts. Most radiologic comparisons are subjective and involve a qualitative visual comparison of the degree of similarity between antemortem and postmortem images, which is insufficient for quantitatively assessing the evidentiary value of an identification. Rather than simply concluding that antemortem and postmortem radiologic comparisons appear the same in the opinion of the examiner, results should be expressed quantitatively. This bolsters conclusions by providing statistical support for the probability of correct identification. Epigenetic trait variation is assessed by a forensic anthropologist during the examination of unknown human skeletal remains and may be useful in establishing positive identification, and/or in providing investigative direction. A key factor in this regard is the frequency of the trait(s) being compared in a given population. The present study utilizes epigenetic trait data from a preceding publication to demonstrate a method of statistically quantifiable positive identification based on epigenetic trait frequencies, ultimately demonstrating the utility of this method in practice. Utilizing a case study approach, the present authors demonstrate the benefits of a combined likelihood approach and propose standards for the presentation of likelihood ratios and verbal equivalent statements, to promote consistency in the reporting of results.