Date of Award
5-2025
Document Type
Thesis
Department
Social Sciences
Committee Chair/Advisor
Katherine Weisensee
Committee Member
Suzanna Tremblay
Committee Member
John Coggeshall
Abstract
Biological anthropologists develop human skeletal biological profiles to identify skeletal remains and build demographic knowledge of past and modern peoples. Biological anthropologists estimate age-at-death, sex, stature, ancestry, and the presence of injury or pathology to develop biological profiles. While biological profiles are at the basis of the work anthropologists perform, there is a lack of research concerning the reproducibility and interobserver agreement rates of the methods used to create biological profiles. Due to the risk of variability stemming from the education, training, and experience of observers, along with variation that can arise from population affinity, examining the interobserver agreement rates of observers using these methods is imperative to determine where variability exists to combat it and strengthen the methods used by anthropologists when developing biological profiles.
This study used human skeletal remains provided by the University of Szeged, Hungary, to examine interobserver agreement on the biological profiles developed by Clemson University students in 2023 and 2024. The aspects of the biological profile analyzed within this research are age-at-death estimation, sex estimation, and the presence of pathology. This research proposed two possible outcomes: either there will be no significant statistical differences between the aspects of the biological profiles created by students in 2023 and 2024, or there will be a significant statistical difference between the aspects of the biological profiles created by students in 2023 and 2024. This study is unique as it solely focuses on biological profiles created by students. Undergraduate and graduate students play an essential role in anthropological research; their roles and contributions are often not considered in this study's analysis. Furthermore, this study adds to a work neglected in anthropological studies. Reproducibility studies are few and far between within the field but they are vital to scientific research. Studies that test interobserver agreement must be conducted to explore where variability exists; only then can it be remedied.
To test the hypothesis proposed, Cohen's Kappa was used to determine the interobserver agreement of observers' estimations of the categorical variables, sex, age categories, and the presence of pathology. T-tests were used to determine the interobserver agreement of the observers' estimations of the continuous variables of numerical age estimations using transition analysis and non-metric methods. The Kappa and T-tests resulted in the findings of moderate agreement of age categories, fair agreement of sex estimation, moderate agreement of the identification of cribra orbitalia, moderate agreement of the identification of osteoarthritis, no significant statistical differences of age estimation using non-metric methods, and no significant statistical differences using transition analysis. The statistical analysis produced resulted in the ability to reject the null hypothesis. Thus, there were no significant statistical differences between the aspects of the biological profiles created by students in 2023 and 2024. While the findings of this study align with the few studies that examine the interobserver agreement of profiles created solely by undergraduate students, the results were typically lower than those conducted using professional anthropologists and graduate students.
While this research found no significant differences between the profiles created by students, there is still much room for improvement as variability exists within every aspect of the biological profiles tested. This research calls for increased reproducibility and repeatability studies within biological anthropology. More research must investigate how observers' educational attainment, training, and experience can affect interobserver variability. Furthermore, this study identifies that an open-science approach could greatly benefit future research. Open access to methodology studies, validity testing, population and observer sample details, and data collection procedures allow for transparency, and future researchers will have an easier time facilitating reproducibility testing. Moreover, allowing students to engage with methodological and reproducibility studies helps them improve their knowledge of the concepts they are introduced to in their anthropology courses. Allowing students unrestricted access to anthropological work equips them with the necessary skills and understanding of the methods they will use in their coursework, produce research, and inspire future professional endeavors.
Additionally, the standardization of methodology and education within biological anthropology must be reinforced. As Galloway and Simmons (1997) and Passalaqua and Pilloud (2020) note, there are discrepancies between the topics, training, coursework, and methods undergraduate and graduate students learn. Standardization is needed in classrooms and anthropological texts to combat these differences that can lead to variability.
Recommended Citation
Bailey, Alleyia, "Interobserver Agreement of Human Osteological Analysis on the Szeged, Hungary Skeletal Collection" (2025). All Theses. 4455.
https://open.clemson.edu/all_theses/4455