Masters Theses

Date of Award


Degree Type


Degree Name

Master of Arts



Major Professor

Dawnie W. Steadman

Committee Members

Benjamin M. Auerbach, Michael W. Kenyhercz


Accurate sex estimation of human skeletal remains is imperative for skeletal biologists, and relies upon the sexual dimorphism between males and females in a population. The degree of dimorphism, and hence the accuracy of sex estimation methods, are known to vary among populations, and while such global patterns have been well studied, the underlying causes of this variation are relatively unclear. Body size—a sexually dimorphic trait that also varies among populations—has previously been shown to affect skeletal morphology, yet whether specific body size parameters, such as stature and body mass, influence the expression of traits used for nonmetric sex estimation has not been previously tested.

To address this problem, this study tests three hypotheses: (1) variation in expression of sexually dimorphic traits will co-vary with body size, (2) stature will be a greater contributor to the sexual dimorphism of specific skeletal traits than body mass, and (3) sexual dimorphism will increase as stature increases.

Over 200 skeletons of varying body size were scored according to three widely-used sex estimation methods for the cranium, pelvis, and humerus. Existing and novel metrics provide more nuanced measures of dimorphism in the same skeletal regions. A suite of parametric and nonparametric statistical tests were conducted to determine the degree of covariation among the skeletal traits, method performance, and body size.

Results show significant correlations exist between dimorphic traits and both body size parameters, with stature exhibiting stronger correlations than body mass. However, the effects of stature on individual traits become diluted when analyzing overall method performance, which remains unaffected by body size. The exception is the pelvic method, which exhibits high classification accuracy for the tallest and shortest males in the sample, with individuals around the mean at the highest risk of misclassification.

Despite the results of the pelvic method, overall there is little effect of body size on morphological sex estimation performance. However, body size does influence metric traits, with stature exerting more influence than body mass. Future research should explore whether body size, particularly stature, has the potential to affect both intra- and interpopulational application of metric sex estimation methods.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."