Adult age estimation from the pubic symphysis and sternal rib: A combined Bayesian revision of the Hartnett methods

Abstract Two of the most commonly used methods for adult age estimation in forensic anthropology utilize the pubic symphysis on the pubic bone and the sternal end of the fourth rib. Currently, there are no age estimation methods that combine visual assessments of the joints of both the pubic symphysis and sternal end of the fourth rib to produce a single statistically derived final age interval. To fill this gap, pubic symphysis and rib scores were modeled separately and together via univariate and bivariate cumulative probit, respectively, with age and sex as predictors, using a sample of pubic bones and sternal fourth rib ends ( N = 481) collected from donors of known age, sex, and population affinity at the Maricopa County Forensic Science Center and Barrow Neurological Institute in Phoenix, Arizona. The bivariate model was more precise and had lower root mean square error (RMSE) while maintaining expected mean accuracy. These results were consistent across the age range, excepting a slight drop in accuracy for younger individuals. Although the rib was more informative, including the pubic symphysis improved the model performance. Sex was a significant predictor of indicator scores, which highlighted the need for having separate tables by sex. Combining and re‐modeling the Hartnett pubic symphysis and sternal rib methods produced one outcome per pair of scores by sex. Simple and easy to use sex‐specific lookup tables are included for practitioner use. Furthermore, old data can be reprocessed with the new lookup tables since the scoring system has not changed.