A Multifactorial Approach to Estimating Geographic Origin of Hispanics Using Cranial and Dental Data

Hispanics are currently the largest minority group in the US and represent a disproportionate number of forensic cases. Current methods available to forensic anthropologists are inadequate for individuals considered Hispanic, producing broad, ambiguous ancestry classifications (e.g. African, European, Asian), which result in high misclassification rates.

In order to reduce the misclassification of individuals considered Hispanic, the proposed research will incorporate multiple data types, cranial and dental metric and non-metric data, from diverse Hispanic reference groups in a multifactorial approach to estimate geographic origin (e.g. Mexican vs. Central American). Previous research suggests that combing various data types can enhance classification power. Therefore, this research will enhance law enforcement investigations by expediting unidentified and missing persons investigations and providing better methods with robust reference data, while meeting the standards and best practices for the field. Cranial and dental data metric and non-metric data will be collected following standards cited in peer-reviewed literature. Population groups will include Latin America (i.e. Mexico, Central America, South America, and the Caribbean), undocumented border crossers (UBCs), as well as European and African groups. European and African groups are included because Hispanics and Latinos represent varying degrees of admixture between Indigenous Americans and these populations due to colonization and the transatlantic slave trade. Therefore, analysis of substructure between these populations will help predict county of origin.

American Black and American White samples will also be included in analysis as previous research has shown Hispanics often misclassify as these groups. Because these data are a mix of both continuous and discrete data types, they will be analyzed using structured equation, random forest modeling, and probabilistic cluster modeling. The end results will be evaluated to determine which combination of variables from the different data types and which statistical analysis provide the best classificatory power. The results of this research will be disseminated through peer-reviewed articles and made available to medicolegal professionals through outreach workshops hosted by the Forensic Anthropology Center at Texas State University.

Note: This project contains a research and/or development component, as defined in applicable law," and complies with Part 200 Uniform Requirements - 2 CFR 200.210(a)(14). ca/ncf