This paper describes the method of osteometric sorting. Osteometric sorting is the formal use of size and shape to sort bones from one another. The method relies heavily on measurement data and statistical models and is designed to maximize objectivity. The goal of this paper is to validate the use of osteometric sorting and to provide case examples of its utility. Selected regression models are also presented for use with osteometric sorting. We advocate this technique as one tool among many in the anthropologist's toolkit for sorting commingled skeletal remains.

Skeletal remains discovered at a construction site in Georgia display classic “butterfly” fractures on several long bones. Although this fracture pattern is usually associated with perimortem trauma, in this case taphonomic indicators demonstrate that they can also occur on dry defleshed bone. Variations in bone color at the fracture sites indicate recent postmortem trauma. Analysis of the directions of force and points of impact indicates that the bones were most likely disarticulated when the trauma occurred.

Osteometric data are of great importance for analytical purposes in the field of forensic anthropology, and it is critical that interobserver concordance is high in order for the results of these analyses to be reliable. Significant interobserver variation of skeletal measurements is cause for concern since it may result in conflicting conclusions. The range of interobserver variation of selected postcranial measurements is addressed. Thirteen standard measurements familiar to most forensic anthropologists were examined, as well as nine nonstandard measurements that were unfamiliar to most participants in the study. Sixty-eight individuals participated in the study, the majority of whom considered themselves to be forensic anthropologists with various levels of experience. In general, interobserver variation of the tested standard postcranial measurements was found to be minimal, with the exception of pubis length and subtrochanteric femur measurements. The difficulties that can lead to measurement error are discussed and possible solutions are recommended.

The primary goal of this research is to examine the overall utility of nonradiographic dental records for the establishment of individual identifications. It was found that even without radiographic lines of comparison, charts and notes that accurately detail a missing individual's antemortem dental condition can be essential for establishing an identification. Based on an analysis of two large datasets, individual dental patterns were determined to be generally unique, or at least very uncommon. Through this type of empirical comparison, it is possible to establish a strong, quantifiable association with a missing individual. The results of this research indicate that a definitive number of points of concordance do not need to be established in dental identification cases. Each case must be assessed individually. The critical factor is to remove subjective judgment calls from dental comparisons. This research has proposed a new method of empirical comparison that allows forensic odontologists to derive objective frequency information regarding the occurrence of specific dental patterns in the general population. The method is similar to that used for mtDNA casework, and a computer program (OdontoSearch) has been developed to make the technique accessible. It was found that even a small number of common dental characteristics may produce a very rare dental pattern, a point that may be counterintuitive to many forensic odontologists.