Semester of Graduation

Spring 2021


Master of Arts (MA)


Geography and Anthropology

Document Type



This study tested the correlation of dimensions of the left seventh rib (L7) to ancestry and sex in adult humans. The first hypothesis, based on Bergmann’s theory, is that Black (African-American) individuals will have smaller L7 dimensions than White individuals (Americans of primarily European descent). The second hypothesis is that males, due to sexual dimorphism, will have larger ribs than females. 299 individuals from the Hamann-Todd Skeletal Collection were used in the study, approximately evenly distributed among sex and ancestry groups. Five variables were examined, including three measurements (length, width, and width point), and two calculations (area and ratio). A Pearson’s correlation tested the precision of the measurements being taken. To further test intraobserver error, 29 individuals measured on the first day of data collection were remeasured on the last day, and those two sets of measurements were analyzed with a paired T-Test. Swarm plots were created to show the distribution of data separated by sex and ancestry, as well as by sex only. Two sample T-Tests were run on all the variables to look for differences in the means with ancestry and sex. Results found significant differences between ancestry groups for both sexes for length, width, width point, and the ratio, with Black individuals smaller than White individuals for all variables except width. Thus, the first hypothesis was only partially supported. The second hypothesis was fully supported; significant differences were found between sexes for all variables, with males being larger. Lastly, non-linear models to estimate ancestry and sex from L7 dimensions were created based on results of the swarm plots, then tested on the main dataset and on a smaller dataset derived from contemporary individuals. From the models, ancestry was correctly estimated in 70% and 91.6% of males, and 65.1% and 66.6% of females, of the test samples, respectively; sex was correctly estimated in ~81% and ~89% of the test samples. Future research should focus on testing inter-observer error, the symmetry of the ribs and applying the models to other population samples.

Committee Chair

Listi, Ginesse A.