Estimation of Sex Through Metric Measurements of the Petrous Portion of the Temporal Bone in Contemporary Populations




Watson, Cristina Chartier

Journal Title

Journal ISSN

Volume Title



The current study seeks to determine if metric measurements of the petrous portion of the temporal bone is an appropriate method of sex estimation for contemporary individuals. Skeletal remains utilized in this study were derived from the William M. Bass Donated Skeletal Collection at The University of Tennessee- Knoxville and the Texas State University-San Marcos donated skeletal collection. Methods utilized for this study were derived from Kalmey and Rathbun (1996) who compiled 9 measurements of the petrous portion and applied them to historic individuals. Intraobserver error was calculated via a 10% approach, comprised of remeasuring 10% of the total sample at the end of data collection and comparing them with the original measurements. Intraobserver error resulted in a Pearson’s correlation coefficient ranging from 0.994 and 0.999, indicating a high level of repeatability and reliability. General Linear Model MANOVA (GLM MANOVA) procedures suggest that age at metric measurements of the petrous portion and sex are statistically significant for measurements D at p= 0.000 and F at p=0.029. Additional GLM MANOVAs indicate that there is a significant relationship between metric measures and age at death by decade intervals for measurement D, and between metric measures and age at death by 2 decade intervals for measurements HI, D, and F. Results show that a GLM MANOVA for metric measures and ancestry were not statistically significant; however, the sample sizes for American Blacks and Hispanics in this study are too small preventing any reliable inferences regarding ancestry and metric measures to be made. Discriminant function analysis resulted in the correct classification of approximately 60% of females, 66% of males with a total cross-validated classification rate of 63%. Stepwise discriminant function analysis selected measurements C and D as the best variables to use together to estimate sex. Utilizing the selected variables created via a linear discriminant function resulted in correct classifications of 58% females, 72% males, and a 66% total correct classification rate. Reapplied to all individuals including those with missing data, the total correct classification increased to 73%. This research shows that the petrous portion can be used as an accurate estimator of sex and can be another method in the arsenal of forensic anthropologists in the absence of other osseous remains.



Biological Anthropology, Sex Estimation, Petrous Portion, Metric Analysis, Forensic Anthropology


Watson, C. C. (2013). <i>Estimation of sex through metric measurements of the petrous portion of the temporal bone in contemporary populations</i> (Unpublished thesis). Texas State University, San Marcos, Texas.


Rights Holder

Rights License

Rights URI