An assessment of sexual dimorphism and sex estimation using cervical dental measurements in a Northwest Coast archeological sample

Abstract

Our goal with this research is to test the consistency of two approaches for the development of sample specific sex prediction methods in fragmentary remains using a Northwest Coast archeological sample from cervical tooth measurements. T-tests and percentage of sexual dimorphism were used to quantify the amount of sexual dimorphism present in the sample (n=42), then, sample specific sex prediction methods were developed using logistic regression formulae where sex was previously assessed from morphological traits of the pelvis (Black, 1978b) and using the sectioning point approach (Albanese et al., 2005). The best performance was achieved by the mesiolingual–distobuccal diameter of the mandibular first molar (LM1MLDB) and the mesio-distal diameter of the mandibular left second molar (LLM2MD), with logistic regression and sectioning point methods showing 86.7% and 85.71% of total consistent sex classifications, respectively. Small sample sizes eliminated the most potentially sexually dimorphic measurements from the analysis and unbalanced sex samples prevented more reliable estimates of consistent sex classification accuracy. However, these findings highlight both the potentials of developing sample specific sex prediction methods in fragmentary archeological samples from cervical tooth measurements, and the pitfalls of relying on small samples when bioarcheologists do not have any other means for sex estimation.