Sociology and Anthropology, North Carolina State University
Saturday All day, Clinch Concourse
Sex estimation from the skeleton is of great importance in both the bioarchaeological and forensic settings. Recently, two computer software programs have emerged that can estimate sex from the os coxa using metric measurements. In the present study, measurements from 58 male and female individuals of both known and unknown sex were entered into FORDISC 3.1 developed by Jantz and Ousley (2005) and DSP (Diagnose Sexuelle Probabiliste) developed by Murail, et al. (2005). Individuals of unknown sex were morphoscopically sexed using the Phenice (1969) method prior to measurements being taken. Per the requirements of the DSP software, the following assessments were performed: all 10 variables, the “best” eight, the “best” four, and the “worst” four. Considering a posterior probability of .95 a success, FORDISC was able to assign sex for 24.14% of the sample, with 67.24% accuracy. When all 10 measurements were input into DSP, 86.21% were able to be sexed (posterior probability (p) ≥ .95), with 96.00% of determinations found to be accurate. Of the three remaining combinations of DSP variable measurements, the “best” four were found to provide the most successful results, with an 89.66% sexing ability rate and 100% accuracy rate. This combination obtained better rates than when all 10 measurements were input. Due to higher sexing and accuracy rates obtained from DSP and its ability to provide more versatility in employable measurements, this software program may offer researchers better sex estimation from the os coxa than FORDISC 3.1, especially if the os coxa is fragmentary.