forensic anthropology research

• Search Menu
• Sign in through your institution
• Advance Articles
• Special Issues
• High-Impact Collection
• Author Guidelines
• Submission Site
• Call for Papers
• Open Access Options
• Self-Archiving Policy
• Why Publish with Us?
• About Forensic Sciences Research
• Journal Metrics
• About the Academy of Forensic Science
• Editorial Board
• Advertising & Corporate Services

Article Contents

Molecular analysis of skeletal evidence, migrant identification, search, detection and recovery, commingling analysis, biomechanics of bone trauma, decomposition research, bone microscopy, isotope analysis, facial imaging, recent advances in forensic anthropology.

• Article contents
• Figures & tables
• Supplementary Data

Douglas H. Ubelaker, Recent Advances in Forensic Anthropology, Forensic Sciences Research , Volume 3, Issue 4, December 2018, Pages 275–277, https://doi.org/10.1080/20961790.2018.1466384

• Permissions Icon Permissions

Forensic anthropology involves diverse applications of anthropological knowledge to medico-legal problems. While the applications are evidence-driven, the available scientific methodology and foundation have developed through decades of research and experience. The roots of this field are anchored in comparative human anatomy but methodology has developed through experimentation, the assemblage of documented collections and databases and thoughtful research design. While forensic anthropology represents a mature scientific field, it continues to evolve and advance through new, innovative global research. Much of this progress is fuelled by issues encountered in casework. The unique evidence and problems presented in forensic cases call for the very best scientific approaches available. Usually, the correct approaches and solutions can be found in the existing scientific literature. However, sometimes the unique issues presented by the casework cannot be addressed adequately with the existing techniques. These situations stimulate forensic anthropologists to seek new solutions through targeted research.

This Special Issue presents research advances in several areas of forensic anthropology that have sustained rapid, recent progress. While our journals continually reveal new information in all aspects of forensic anthropology, several areas of investigation have registered particularly strong academic interest featuring innovative research.

Successful recovery and analysis of DNA has dramatically affected many areas of forensic science. In the field of forensic anthropology, molecular analysis can yield highly accurate information regarding the sex of the individual represented and provide positive identification [ 1 ]. Molecular approaches also can contribute to ancestry evaluation and species recognition. The use of DNA for positive identification has had a major impact on the practice of forensic anthropology and related fields of forensic science.

While the merits and contributions of DNA analysis are profound, many related issues express the need for new, innovative research and technological development. Frequently, evidence submitted for forensic anthropological analysis is not in pristine condition. In many cases, recovered remains are incomplete and/or extremely degraded due to criminal activity and/or taphonomic factors. Some site investigations produce only small fragments where even species is not apparent. Decisions need to be made regarding what areas of bone or tooth should be examined. Since DNA analysis is an expensive and destructive process, these decisions are critical and can affect the outcome of the case. Of course, decisions regarding the type of DNA analysis also are critical and largely driven by both the availability of the antemortem information and the nature of the evidence. Experimentation and casework experience have greatly improved approaches to these issues.

Deaths related to the global movement of undocumented people across national borders present major forensic challenges. Even within countries, identification of citizens can be difficult with incomplete evidence and lack of information regarding missing persons. These problems are greatly exacerbated when different countries are involved and the international movement of the person represented is not registered officially. Such cases call for extraordinary investigation, thoughtful forensic analysis and international communication. These efforts can strain the available local resources and often fall short of positive identification.

Recent years have witnessed remarkable efforts to address the identification of deceased, undocumented border crossers. These initiatives have involved international cooperation, careful exhumation procedures, comprehensive anthropological analysis and new techniques such as isotope analysis to identify the likely regions/countries of origin.

The entire process of forensic anthropological investigation begins with the procedures of search, detection and recovery. Improper or inadequate detection and recovery of human remains can compromise the downstream analysis and interpretation. While the traditional techniques of surface survey and excavation continue to be needed, new approaches, especially those using advanced technology offer significant advances.

Search procedures can be especially challenging when only very general information is available regarding the likely location of human remains. Topographic features can present limitations, especially with dense vegetation and other ground cover. Investigations of humanitarian and human rights issues can present special search and recovery challenges when information suggests that wells, cisterns, sewer systems, mass graves or disposal in water were involved. Confronted with these problems, researchers have devised innovative new approaches to improve the probability of success.

Secondary deposits of human remains or those that have sustained significant disturbance involve loss of normal bone articulation patterns. When multiple individuals are involved, the resulting commingling presents challenges to determine the number of persons represented and to assemble remains of individuals for analysis, identification and return to families. Traditional approaches to commingling problems have involved sorting by the type and side (left or right) of bone, age at death, bone size and maturation, sex and pathological conditions. In some skeletal assemblages, taphonomic indicators can be helpful as well.

Once obvious sorting has been completed, questions persist regarding bone morphology related to individuals. Could a robust femur relate to a robust humerus and represent one individual? Recent advances in commingling analysis address this issue. New databases and computerized techniques establish the probabilities that different bones could relate to the same individual. Applications refine the determination of the number of individuals represented and facilitate analysis aimed at identification.

A primary function of anthropological analysis relates to the interpretation of bone trauma. Anthropologists must differentiate the skeletal alterations representing perimortem trauma from those relating to antemortem injury, developmental features or postmortem and taphonomic factors. Assessment of the biomechanical factors involved plays a key role in any interpretation. Knowledge of biomechanical principles is required to explain fracture patterns and other alterations likely related to perimortem trauma. Interpretation of bone trauma can be challenging. Such challenges have led to greater understanding of the principles involved and experimental work designed to improve interpretation.

Major new initiatives in forensic anthropology have focused on decomposition research. Experiments involving both humans and non-human animals have revealed great detail about the process and variation of soft tissue decomposition and hard tissue alteration. In general, such research has elucidated the many factors that influence both the nature and timing of the decomposition process. Clearly temperature and location (surface, in-ground, aquatic, etc.) have long been regarded as key factors. Research has also indicated that soil conditions, moisture, body composition, body condition, presence of clothing or enclosures, funerary treatment and many other factors can influence the process. Such information is needed to properly assess time since death (post-mortem interval) and post-mortem events related to criminal activity.

In 1965, Ellis R. Kerley [ 2 ] published a technique that allowed age at death to be estimated from microscopic examination of features in human compact bone from the femur, tibia and fibula. Kerley's procedure involved the examination of primary osteons, secondary osteons, osteon fragments and the extent of remaining circumferential lamellar bone. This approach gained recognition due to its reported accuracy and the fundamental processes of bone formation and remodelling that it expressed. Since 1965, the technique has undergone many revisions and expansions for application to other bones of the skeleton. Research also has revealed how bone microscopic examination can provide useful information on many issues of forensic anthropological analysis.

For decades, analysis of elemental stable isotopes has offered key anthropological information related to diet. Stable carbon isotopes recovered from human tissues have revealed if diet focused on plants with a C 3 photosynthetic pathway or a C 4 pathway and the herbivores that fed upon them. Analysis of nitrogen isotopes provides insight into the trophic level of human diet. In anthropological studies of ancient populations, such information is crucial to interpretations of dietary and horticultural practices.

Recently, researchers have applied the concepts of isotopic analysis to examine the geographical origin of human remains. When unidentified human remains are recovered in forensic contexts, investigators question if they represent someone who lived in the area of recovery or from somewhere else. This question is especially relevant in cases involving terrorism and unidentified possible migrants. Using a battery of stable isotope analyses, researchers can determine if the isotopic signatures from the unknown match local baseline data. If not, attempts can be made to determine from what geographic area the unknown originated. This exciting new area of forensic science investigation depends on the assemblage of baseline data from appropriate geographic regions.

Forensic anthropologists relate to issues of facial imaging in facial approximation, craniofacial photographic superimposition and interpretations of surveillance images. Facial approximation refers to the process of estimating the living facial image of a person from the evidence presented by a recovered skull. This technique is used to reach out to the public for leads in missing persons that could culminate in identification using other methods.

Craniofacial photographic superimposition involves comparing a facial photograph of a missing person with a recovered skull. This technique is used primarily to exclude when photographs are available of a missing person thought perhaps to be represented by the recovered remains.

Recent research has focused on enhanced use of computers and related technology, as well as targeted efforts to clarify the relationship between soft and hard tissues. Facial approximation continues to represent a blend of art and science; however, recent advances have strengthened the scientific foundation.

Articles in this Special Issue of Forensic Sciences Research focus on overviews of the published literature on these topics. They also share results from the latest innovative research on these key areas of forensic anthropology applications.

Baker   L . Biomolecular applications . In: Blau   S   Ubelaker   DH , editors.   Handbook of forensic anthropology and archaeology . 2nd. ed. New York: Routledge ; 2016 . p. 416 – 429 .

Google Scholar

Google Preview

Kerley   ER . The microscopic determination of age in human bone . Am J Phys Anth . 1965 ; 23 : 149 – 163 .

Email alerts

Citing articles via.

• Advertising and Corporate Services
• Journals Career Network

Affiliations

• Online ISSN 2471-1411
• Print ISSN 2096-1790
• Copyright © 2024 Academy of Forensic Science
• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Institutional account management
• Rights and permissions
• Get help with access
• Accessibility
• Advertising
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Forensic Anthropology

What do forensic anthropologists and detectives have in common.

Forensic anthropology is a special sub-field of physical anthropology (the study of human remains) that involves applying skeletal analysis and techniques in archaeology to solving criminal cases. When human remains or a suspected burial are found,  forensic anthropologists are called upon to gather information from the bones and their recovery context to determine who died, how they died, and how long ago they died.  Forensic anthropologists specialize in analyzing hard tissues such as bones. With their training in archaeology, they are also knowledgeable about excavating buried remains and meticulously recording the evidence. 

Reading a Skeleton

A forensic anthropologist can read the evidence in a skeleton like you read a book. The techniques they use to answer questions in criminal cases can be applied to skeletons of any age, modern or ancient. The stages of growth and development in bones and teeth provide information about whether the remains represent a child or adult. The shape of pelvic bones provides the best evidence for the sex of the person. Abnormal changes in the shape, size and density of bones can indicate disease or trauma. Bones marked by perimortem injuries, such as unhealed fractures, bullet holes, or cuts, can reveal cause of death. The trained anthropologist is also able to identify skeletal clues of ancestry. Even certain activities, diet, and ways of life are reflected in bones and teeth.

Analyzing Human Remains

Anthropologists at the Smithsonian’s National Museum of Natural History have been called upon to analyze human remains for over a century. The remains may represent victims of violence or natural disasters. In these cases Smithsonian anthropologists work with the FBI , State Department, and other law enforcement agencies to identify the individuals and solve crimes. They also conduct research on historic and prehistoric human remains to learn more about people from the past. As Smithsonian forensic anthropologist Kari Bruwelheide says, "The bones are like a time capsule."

Smithsonian anthropologist Dr. Douglas Owsley , examining a skeleton from historic Jamestown , discovered evidence of chops to the skull from an axe or other sharp bladed, implement. Knife cuts were also observed on the bone. Along with other information such as biological indicators and discovery location of the remains, Dr. Owsley concluded that a 14-year-old girl had been cannibalized after she died. His discovery supported other historic data that the colonists of Jamestown suffered severe starvation during the harsh winter of 1609-1610.

Techniques: Leaving No Bone Unturned

Anthropologists at the National Museum of Natural History use a variety of techniques to analyze human remains and record their observations. For example, the bones are typically photographed and X-rayed. Some remains may undergo CT scanning or be examined with high-powered microscopes. These techniques provide detailed information about remains without altering them while providing a visual record. DNA analysis may be used to help establish identity. This type of testing is most often used in modern forensic case work, but mitochondrial DNA in bones and teeth can be used to confirm relationships of old remains with deceased or living descendants. Other chemical analyses, such as those involving isotopes, can provide information about the age of bones and a person’s diet.

The data gathered is studied and combined to draw conclusions about the deceased individual. For a modern case, photos of the skull may be superimposed on photos of missing people to look for consistencies between the bone and fleshed form. Even in cases where no photos exist, the face can be reconstructed based on the underlying bone structure and known standards of facial tissue thicknesses. For example, using facial reconstruction, Smithsonian forensic anthropologist Dr. David Hunt was able to bring about correct identification of the remains of a child found near Las Vegas. Owsley and Bruwelheide were able to help rebuild the likeness of the girl from Jamestown .

Collections of Bones

Comparing found remains to other human skeletons is essential for many analyses. The National Museum of Natural History has one of the world's largest Biological Anthropology collections , with over 30,000 sets of human remains representing populations from around the world. Many of the skeletons have associated age, sex, ancestry, and cause of death data. Individual remains with known biological information are especially valuable references.  Forensic anthropologists have used these skeletons to develop standards for determining sex, age and ancestry in unknown remains. The bones and teeth are also used as comparative materials in cases where interpretation of certain features is difficult. They are also used to train students who are the next generation of biological anthropologists. Skeletal reference series may also be used to document trends in health and population structures over time. Smithsonian Curator Dr. Douglas Ubelaker , looking at a range of skulls from 16th-20th century Spain and Portugal, found that women's faces got larger over time.

Reconstructing the Past

The study of historic human remains by biological anthropologists at the Smithsonian has led to discoveries that are changing our view of the past and how we investigate it. The work of Dr. Owsley and Kari Bruwelheide has helped create a better picture of how people lived and died in colonial America. For example, even a wealthy woman, the wife of the governor of Maryland's first English colony, St. Mary's City, suffered from limited medical care for a fractured thigh bone. The sorts of treatments that would be used today (traction and screws), were not options at the time. Available treatments, such as medicine containing arsenic, may have made conditions worse. Chemical testing of this woman's preserved hair show ingestion of this toxin with increasing dosage closer to death. 

Whether used to better understand modern or historic remains, the tools and techniques of forensic anthropology give the living a window into the lives of the dead.

• Smithsonian Institution
• Terms of Use
• Privacy Policy
• Host an Event

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Research Article

Investigating identification disparities in forensic anthropology casework

Roles Conceptualization, Formal analysis, Methodology, Project administration, Writing – original draft, Writing – review & editing

* E-mail: [email protected]

Affiliations Department of Anthropology, University of Illinois at Champaign-Urbana, Urbana, Illinois, United States of America, Carl R. Woese Institute for Genomic Biology, University of Illinois at Champaign-Urbana, Urbana, Illinois, United States of America

Roles Formal analysis, Methodology, Visualization, Writing – original draft, Writing – review & editing

Affiliations Department of Health and Exercise Sciences, Truman State University, Kirksville, Missouri, United States of America, Forensic Science Program, George Mason University, Fairfax, Virginiai, United States of America

Affiliation Department of Anthropology, California State University Fresno, Fresno, California, United States of America

Roles Data curation, Writing – original draft, Writing – review & editing

Affiliation Forensic Anthropology Unit, University of North Texas Center for Human Identification, Fort Worth, Texas, United States of America

Affiliation Trace Evidence Unit, Laboratory Division, Federal Bureau of Investigation, Quantico, Virginia, United States of America

Roles Conceptualization, Writing – review & editing

Affiliation Anthropology and Sociology Department, Western Carolina University, Cullowhee, North Carolina, United States of America

Roles Writing – original draft, Writing – review & editing

Affiliation New York City Office of Chief Medical Examiner, New York City, New York, United States of America

• Cris Hughes, 
• An-Di Yim, 
• Chelsey Juarez, 
• John Servello, 
• Richard Thomas, 
• Nicholas Passalacqua, 
• Angela Soler

• Published: November 1, 2023
• https://doi.org/10.1371/journal.pone.0290302
• Reader Comments

Forensic anthropology is shifting to reflect on the impact of its practices within the criminal justice context in important ways. Here, we contribute to this essential work by examining how decedent demographics as well as estimations of biological profile components are related to identification trends in forensic anthropology cases. The study uses data from more than 1,200 identified and unidentified forensic anthropology cases from three agencies (together representing a nation-wide sample). We found the following: i) multivariate analyses indicated that decedent sex, age, and race and/or ethnicity are not related to case identification rates in the pooled United States sample, ii) when identification rate differences do occur, they appear to be smaller effects, more agency-specific, and/or related to the context of a particular agency, iii) for the agency-specific sample with available data, there was no consistent evidence for a discrepancy in the duration of an identification investigation based on a decedent’s sex, age, or race and/or ethnicity, iv) forensic anthropological estimations of sex, age, and ancestry can improve the odds of identification for decedents, although these are small effects, and v) reporting an ancestry estimation does not appear to impact decedent race representation among resolved unidentified person cases. Although previous studies have identified demographic discrepancies in other areas of the criminal justice system, the results presented here suggest that decedent demographic estimation practices by forensic anthropologists in general do not appear to be related to discrepancies in identification trends, but more research is needed to examine whether these findings hold. Contextual factors and practices specific to each investigative agency likely contribute to identification trends.

Citation: Hughes C, Yim A-D, Juarez C, Servello J, Thomas R, Passalacqua N, et al. (2023) Investigating identification disparities in forensic anthropology casework. PLoS ONE 18(11): e0290302. https://doi.org/10.1371/journal.pone.0290302

Editor: Roberto Scendoni, University of Macerata: Universita degli Studi di Macerata, ITALY

Received: February 15, 2023; Accepted: August 5, 2023; Published: November 1, 2023

This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.

Data Availability: Regarding data sharing, deidentified data used in this study cannot be openly shared in the journal for legal reasons, as the underlying data is law enforcement-sensitive. Working with this data required permission from the three contributing forensic agencies. Data underlying the results presented here may be requested from the FBI ( https://www.fbi.gov/investigate/how-we-investigate/science-and-technology ; [email protected] ), UNT ( https://www.unthsc.edu/center-for-human-identification/ ; [email protected] ), and NYC OCME ( https://www.nyc.gov/site/ocme/index.page ; [email protected] ). Restrictions may apply to the availability of these data, please contact the respective agency to inquire.

Funding: The author(s) received no specific funding for this work.

Competing interests: authors Soler, Servello, and Thomas each work in an agency from which data was collected and analyzed for this study. This does not alter our adherence to PLOS ONE policies on sharing data and materials.

Introduction

With the National Academy of Science’s report, Strengthening Forensic Science in the United States : A Path Forward over a decade in the rearview mirror, research in forensic anthropology has consistently and robustly addressed measures for assessing the accuracy and reliability of approaches to forensic casework. While the standards laid out in that report will continue to inform methodological approaches and research in forensic anthropology for years to come, recent discussions and work in the discipline have highlighted the need to give equal consideration to researching the contextual impacts of the practice of forensic anthropology—in other words, discovering and coming to terms with both positive and deleterious downstream effects of forensic anthropology practice in the investigation context, as well as the broader social context.

In the summer of 2020, nationwide demands for radical change to the criminal justice system were ignited as a result of the high profile (but not isolated) incidents of a series of police killings of Black individuals in the United States. Within the discipline of forensic anthropology, researchers simultaneously called for the critical examination of ancestry estimation as it relates to concepts and applications of race in research and unidentified persons casework. Specifically, Bethard and DiGangi [ 1 ] asserted that the practice of ancestry estimation of unidentified human skeletal remains by forensic anthropologists is doing harm to communities of color by reifying racialized science, hindering identifications in casework, and applying methods with skeletal data for which we lack evolutionary understanding. In response, Stull and colleagues [ 2 ] countered that ancestry estimation plays a role in repatriation and determination of medicolegal significance and cautioned against abandoning ancestry estimation without consultation with both stakeholders and anthropologists in the discipline. While providing different viewpoints, both Bethard and DiGangi [ 1 ] and Stull et al. [ 2 ] did align regarding the need for data-driven research to investigate each others’ proposed perspectives. This conversation garnered national attention and laid the groundwork for recent critical discussions and research on issues of diversity and inclusion [ 3 – 6 ], the perception and framing of race and ancestry [ 7 – 12 ], as well as gender and biological sex [ 13 – 15 ], and reexamination and improvement of the language used in our research [ 16 , 17 ]. As forensic anthropology works to critically engage with concepts beyond the science itself and associated methodological approaches, it is essential to recognize that this discipline is part of the larger community of death investigation, for which similar conversations are rarely happening (with a few exceptions, see [ 18 , 19 ]), yet are equally necessary. Forensic anthropologists’ work at self-criticality has implications not only for our own practice, but also has the potential to impact and improve the practice of the greater medicolegal death investigation community [ 20 ].

One goal of forensic anthropological analyses in medicolegal death investigations is to produce information from an unidentified person’s skeletal remains that narrows down the pool of potential missing persons from which this person may have originated, thus producing an investigative lead that may assist in the successful identification of the person. Forensic anthropologists engage with socially-constructed concepts of race and ethnicity because these are employed as criteria used to narrow the pool of potential matches within the missing and unidentified person investigation context in the U.S. [ 2 , 21 ]. Furthermore, tracking decedent race and ethnicity allows researchers to provide insight into decedent trends. Recent studies have provided forensic anthropology with a clearer sense of the demographics of decedents common in forensic anthropology cases where the decedents were successfully identified [ 22 , 23 ]. Notably, research indicated that Black and Hispanic persons were disproportionately overrepresented in forensic anthropology casework when compared to U.S. census data population proportions. While insight into demographic representation is an important first step, it has been difficult to comprehensively track investigative trends that may produce and/or result in investigative outcome disparities (e.g. successful identification rates) across demographics for forensic anthropology casework. Researching investigative trends, strategies, hurdles, and biases—all of which may differ at an agency level and can influence the investigative success of an unidentified persons case—include myriad factors for which readily available, standardized data do not exist. However, given that these factors influence investigative outcomes (e.g. duration of an investigation to completion; successfully making an identification), exploring whether there are trends and/or disparities in these outcomes is a logical first step. Here, we explore such investigative outcomes within the context of decedent demographic factors of sex, age, and race/ethnicity, and the information provided by the forensic anthropologist, using available data on casework success and investigation duration. Specifically, this study addresses the following:

• Aim 1. For forensic anthropology cases , do investigation trends (e . g . identification success rate; duration of an investigation) differ across demographic backgrounds (e . g . sex , age , race/ethnicity) ? Based on the extensive bias and structural violence and vulnerabilities documented in the criminal justice system and forensic investigations (discussed in depth below), we hypothesize that Black, Indigenous, and People of Color (BIPOC) cases may have lower success rates or longer investigative durations for identifying the skeletal remains as a missing person. In the present study, our first aim is to test this hypothesis using decedent demographic data from over 1,200 U.S. forensic anthropology cases of unidentified persons (both resolved and open).
• Aim 2. Is the decedent information included in forensic anthropology reports (e . g ., estimations of decedent ancestry , sex , and age) related to the (de)prioritization of particular individuals over others in regard to identification efforts . This aim corresponds to the hypotheses proposed by Bethard and DiGangi [ 1 ] that information related to decedent race/ethnicity (e.g. forensic anthropologists’ ancestry estimations) of an unidentified person may provide grounds for “racial bias on the part of investigators” which “may hinder identification efforts.” Indeed, they propose that the very information provided by forensic anthropologists may be a driving factor for inadequate efforts by investigators to resolve cases of non-White decedents [ 1 ]. While data that directly correspond to prioritization by investigators are unavailable in the present study, proxy data are employed, including the duration of an investigation as well as success rates of identifications. Given that Bethard and DiGangi [ 1 ] hypothesize that investigations may be hindered when relevant demographic information is provided to investigators, investigative success and duration are proxies that represent the comprehensive investigation. Importantly, investigation hinderances and/or deprioritizations would be included in the factors influencing investigative success and duration, and if sufficient to influence the investigation would be observable in the present analysis. Bethard and DiGangi’s hypothesis is explored in several ways, using data on ancestry estimation availability, and when estimated, what those corresponded to in terms of race/ethnicity. The analysis was further expanded to include both age and sex estimations, to comprehensively explore how forensic anthropologists’ estimations of demographic data influence an investigation.
• Aim 3 . Does providing information related to the decedent’s age , sex , and/or race/ethnicity influence the odds of an identification ? For this research aim, the hypothesis is based on Bethard and DiGangi [ 1 ], who challenge the assumption that ancestry estimations are critical for contributing to successful investigations of unidentified persons. We extend the testing of this assumption beyond ancestry estimations to include estimations of age and sex. We compare the identification rate for cases with and without these estimations.

Documenting decedent demographic and investigative trends in United States forensic casework allows for a better understanding of how institutional (i.e., United States Criminal Justice System) and societal factors may be related to, or directly affecting the time to identification and the overall success of identification for various populations within the United States. Thus, documenting demographic information and analyzing these data for trends can in turn have important implications for improving approaches to identification investigations [ 24 – 26 ].

It is worth noting that forensic anthropological analyses typically represent a small component of identification investigations, and demographic and investigative trends potentially vary depending on which aspect of the forensic setting one examines (e.g., all medicolegal death investigations versus forensic anthropology cases). However, the goal of this study is to explore whether there is a relationship between identification status of decedents who receive an anthropological analysis and demographic factors (including sex, age, and race and/or ethnicity) to infer whether any differences in case identifications exist at a regional and/or national scale.

Forensic anthropology’s discourse on ancestry estimation and race

The present focus on examining whether investigation and identification trends differ among decedent demographic groups stems from the much broader discourse within the discipline of forensic anthropology. The broader discourse was borne out of the initial work of Bethard and DiGangi [ 1 ], in which the authors called on the U.S. forensic anthropology community to stop conducting ancestry estimations in forensic anthropology casework for three main reasons. First, they insisted forensic anthropologists have not considered whether ancestry estimations might hinder identification efforts, related to the entrenched (conscious and unconscious) racial biases within the criminal justice system at large. That is, for example, unknown human remains whose ancestry estimation is consistent with the decedent’s race being Black or African American (but potentially extends to BIPOC individuals as a whole) may receive less investigative attention/prioritization than those decedents considered White. Secondly, the authors proposed a framework in which the practice of ancestry estimation reifies race as biology and perpetuates typological notions of race. Essentially, they argued that the act of ancestry estimation itself, the categorizing of humans based on phenotypes is in and of itself the “reliving” of a practice deeply imbedded in a racist history which supported not only slavery, the divestment of Native Peoples from their land, and Jim Crow, but is allowed to continue because of the lack of a reckoning with biological anthropology’s racialized past [ 27 , 28 ]. Thirdly, Bethard and DiGangi [ 1 ] demand a cessation to the use of macromorphoscopic traits in ancestry estimation without a full grasp of the heritability of these traits.

In response to the initial publication by Bethard and DiGangi (1), scholars and practitioners have begun to open up the conversations to an array of relevant findings. Adams and Pilloud (7) explored instances of misappropriation of psychology and biological anthropology research by white nationalists to justify their racist beliefs, and documented ways in which biological anthropology research was used to argue for the existence of racial typology and urged researchers to consider the broader implications of their research beyond their discipline and intended application. Other recent works [ 10 , 11 , 29 – 32 ] recommend moving towards a practice of estimating population affinity, which advances beyond continental allocations often associated with ancestry estimations, and incorporates a comprehensive understanding of population histories as well as ongoing dynamics (sociopolitical, economic) that can influence local patterns of biological variation. In turn, this approach considers how categories and terms used to describe the reference samples represent social meaning relevant to local communities [ 17 , 30 , 31 ]. Several studies propose that current methodological approaches and reference sample representation may be impacting the accuracy rates of ancestry estimation, and suggest that increasing diversity of reference samples would better serve the forensic anthropological casework demographics [ 22 , 23 ]. Furthermore, Go and colleagues [ 33 ] documented how those conducting ancestry-related research, as well as the skeletal samples used, are largely WEIRD (Western, educated, industrialized, rich, and democratic). The authors rightly pointed out that this lack of diversity among practitioners and samples reifies harm and “informs the history, values, and practices of forensic anthropology on a global scale” (Go et al., [ 33 ], p. 156).

Established demographic trends in U.S. investigations

The previously-reviewed body of work provides data points forensic anthropologists can use to inform their decision (both collectively and individually) of whether and how they will practice ancestry or population affinity estimation moving forward. The current study provides another “data point” for consideration in the larger ongoing conversation regarding the contextual impacts of our practice. Furthermore, this study intentionally expands the inquiries beyond ancestry estimation, to incorporate estimations of sex and age, as research has established that other demographic factors in addition to or in tandem with race/ethnicity are related to investigation and criminal justice disparities. Bethard and DiGangi [ 1 ] and DiGangi and Bethard [ 34 ] referenced sociological work on “missing White woman syndrome” as an example, which highlights how certain cases (those involving White women) are prioritized in social media and news outlets over others [ 35 ] and therefore have greater odds of resolution. Other studies have explored the extent to which BIPOC individuals and/or members of the LGBTQ+ community are negatively impacted by the criminal justice system in a variety of ways, from police engagements, to investigative outcomes, to sentencing and parole results [ 36 – 44 ]. Tallman et al. [ 15 ] surveyed 128 forensic anthropologists and found 28.9% had encountered transgender individuals as part of their casework and called for greater consideration of queer theory in forensic anthropology research and practice.

Within the medicolegal death investigation system, Goliath and Cosgriff-Hernandez [ 39 ] comprehensively reviewed racial disparities and biases, including the criminal justice system policies that have led to a disproportionate number of BIPOC individuals and adults coming from vulnerable circumstances represented in the system. Studies have suggested that investigations of missing Black individuals remain unresolved more often, and when resolutions do occur, take longer than cases involving White individuals [ 45 , 46 ]. This statistic is particularly concerning as we know that Black individuals make up a disproportionate number of missing and unidentified cases in the United States at 33%, yet only comprise approximately 13% of the United States’ population [ 47 ]. Related to this discrepancy, it has been argued that differential treatment of various demographic groups has ultimately influenced who is comfortable and/or willing to engage with law enforcement regarding investigations, including investigations of unidentified remains and missing persons [ 48 – 53 ].

Based on the works reviewed here, scientific investigation is warranted regarding whether demographic factors are related to identification status (identified or not) and the rate (i.e., death investigation duration) at which a case is resolved. If such disparities exist, the reasons can be multifactorial, such as investigative practices, practitioner choices, public participation and/or trust in investigations, and structural vulnerabilities. Collectively, the works cited above suggest that there is much more to understand regarding decedent demographic trends related to forensic anthropology practice. By studying the relationship of decedent demographics and identification trends for forensic anthropology casework, we can begin to understand the complex roles that forensic anthropologists’ estimation of a decedent’s biological profile (including sex, age, and ancestry) may have on the identification process. The results of this study can help inform practitioner strategies and policies for mitigating investigation biases.

Addressing analytical limitations: Data and interpretations.

It is important to note that if demographic differences in case identification trends are found in this study, it does not provide evidence for a direct correlation to overt prioritization of one case demographic over another; however, it would provide a foundation and direction for further study into what may be driving such differences. Concomitantly, if demographic differences in case identification trends are not found, it does not provide evidence that medicolegal death investigations are wholly unbiased. Because case investigations (and their success or failure) depend on myriad factors, this study is only a first step in exploring identification trends related to decedent demographics in United States forensic anthropology casework.

It is important to set expectations of studies with case datasets like this up front, and it is helpful to consider comparable roadblocks highlighted in criminal justice research. Knox and Mummolo [ 54 ] examined why research on whether and to what degree police behavior is racially biased was not as forthcoming and conclusive as one might hope, given the heightened attention by scholars in recent years. They emphasize that data constraints/limitations only allow the study of isolated aspects of police–civilian encounters, and thus do not fully represent the myriad of choices, interactions, standard operating procedures, and other factors related to case contexts. Indeed, the data being relied upon by researchers are largely generated by police for police administrative purposes, ultimately producing a fragmentary data source that can lead to incomplete and disparate results when analyzed, given that the original data-generation was not geared towards answering the kind of questions that are now in the forefront of many social scientists’ minds. Relating to the present study, we acknowledge that medicolegal death investigation data of forensic anthropology cases largely originates with the same purposes, for case management and tracking, and is bound by all of the nuanced choices of what to track (and what not to track) and how to track it. If disparities in identification trends exist, then it would be ideal to explore elements of an investigative context that could produce such disparities, such as identification modality, availability of missing person reports/antemortem records/family reference DNA samples, and the quality of evidentiary data, and more. Here, we initiate the work for establishing whether investigative disparities for forensic anthropology cases exist, but the exhaustive factors that may be producing disparities are not readily available in this study. For this initial study, we focus on decedent sex, age, race, as well as agency, case year, and geographic location.

Materials and methods

The forensic anthropology unidentified persons cases used in this study were drawn from three agencies’ databases: the Federal Bureau of Investigation Laboratory, the New York City Office of Chief Medical Examiner (NYC OCME), and the University of North Texas Center for Human Identification (UNTCHI) Forensic Anthropology Unit (FAU). Table 1 provides data for the three agencies regarding sample sizes, the case year ranges, and the percent of individuals identified in each agency sample. Both resolved (i.e. person is identified) and unresolved (i.e. persons’ remains unidentified to date) cases were included. Below is a brief description of any additional considerations for each agency dataset. To provide consistency in casework across the three agencies, cases were included in the sample if they met all of the following criteria: the remains were decomposed beyond the fresh stages, estimation of the biological profile was completed, no tentative, presumed, or positive identification was known upon case intake, and cases were determined to be of medicolegal significance. Tentative identification data were inferred from the case investigation notes or directly reported from the investigating agency. For resolved cases, many had comprehensive information on the individuals’ demographic data for sex, age, and race and/or ethnicity, however some were missing information which is why the sample sizes vary per analysis performed.

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

https://doi.org/10.1371/journal.pone.0290302.t001

FBI case dataset.

The Federal Bureau of Investigation (FBI) Laboratory performs a broad range of forensic examinations on skeletal remains and maintains detailed files for a large number of resolved and unresolved unidentified decedent cases submitted by jurisdictions nationwide. These case files often contain an anthropological biological profile assessment, including estimates of age, sex, ancestry and stature. For those cases with successful positive identification of the decedents via DNA analysis or other investigative means, the biological profile characteristics of the decedents are known and documented. These cases contain anthropological examinations conducted both by anthropologists at the FBI Laboratory and outside of the FBI Laboratory, providing a variety of analyses and data sources. Overall, 525 cases (with most anthropological examinations dating from 2000 or later) were selected. In general, cases with older anthropological examinations were submitted to the Laboratory for DNA analysis, which was not available at the time of recovery, or for facial approximation as part of cold case initiatives carried out by local medicolegal authorities. Reported sample variables used in the present study include: case recovery and analysis years; case identification status (identified or not identified); forensic anthropological estimates of sex, age, and ancestry for the decedents; known decedent demographic data (sex, age, race); and the National Missing and Unidentified System’s (NamUs) region code for case origin. Of note for this study, a key feature of the FBI dataset is that the vast majority of the forensic anthropology cases were investigated by an outside agency with jurisdiction over the case, not the FBI. Therefore, the FBI’s involvement in these cases is limited to the services requested from the FBI Laboratory by the outside agency and does not include ongoing investigation of these cases. In this study we will refer to the FBI agency dataset, but it is necessary to recognize that the FBI’s dataset represents cases worked by forensic anthropologists from an array of agencies and that the FBI does not have jurisdiction over these cases.

NYC case dataset.

This portion of the study dataset is comprised of resolved and unresolved cases (n = 368) from the New York City Office of Chief Medical Examiner (NYC OCME). For those cases with positive identification of the remains, the biological profile characteristics of the decedents are known and documented. These cases span several decades, with the majority of cases with forensic anthropological analyses carried out from 2004 onward, but some of the cases prior to this have been retroactively examined by forensic anthropologists at the agency. The majority of cases in this sample are from the New York City jurisdiction, but some are from other agencies in the surrounding areas within New York State. Thus, the originating agency for the majority of the NYC dataset is the same as the agency performing the forensic analysis and investigation of the cases, which is very different from the situations described for UNT and FBI. Reported sample variables used in the present study included: case recovery and analysis years; case identification status (identified or not identified); forensic anthropological estimates of sex, age, and ancestry for the decedents; known decedent demographic data (sex, age, race); and the NamUs region code for case origin.

UNT case dataset.

This portion of the study dataset consists of resolved and unresolved cases (n = 361) examined at the University of North Texas Center for Human Identification (UNTCHI) Forensic Anthropology Unit (FAU) from 2008 to 2020. Remains examined at the FAU were submitted by investigative agencies nationwide. Cases were recovered between 1960 and 2020, the majority of which were found between 2012–2018. For those cases with positive identification of the remains via DNA analysis or other investigative means, the biological profile characteristics of the decedents are known and documented. Additionally, cases with positive molecular associations (i.e., “DNA identifications”) were coded for presence/absence of available reference profiles (Family Reference Samples or CODIS-Convicted Offender Index). Reported sample variables used in the present study included: case recovery and analysis years; case identification status (identified or not identified); identification year; forensic anthropological estimates of sex, age, and ancestry for the decedents; known decedent demographic data (sex, age, race); and the NamUs region code for case origin. A key feature of the UNT dataset is that it does not have jurisdiction over the cases they examine because they are from other investigative agencies that are outsourcing the forensic anthropology analysis to UNT. Thus, UNT’s involvement in these cases is limited to the services requested by the originating entity and is not directly involved in the ongoing investigation of these cases. In this study we will refer to the UNT agency dataset, but it is necessary to recognize that UNT’s dataset represents cases worked by forensic anthropologists at UNT, however UNT is not the originating agency for any of these cases.

Case data standardization.

In the dataset, there is both known and estimated demographic information about the decedents. Known demographic information is derived from resolved cases when the decedent is positively identified, and includes sex, age and race/ethnicity often derived from legal or medical documentation/records. Not all resolved cases (i.e. identified cases) contain sex, age and race/ethnicity information, as these may not be known or available to the agency at the time of case resolution. Estimated demographic information is derived from the forensic anthropological analysis of the skeletal remains, and includes the estimation of biological sex, age, and ancestry. Not all resolved or unresolved cases have estimated demographic data, due to limitations with the analyses such as damaged or missing skeletal elements used by forensic anthropologists for these estimations. Both known and estimated demographic data are used separately or in conjunction as appropriate per study analysis. These are described in detail in the Supporting Information.

All but multivariate analyses were performed using JMP ® Pro 16.1.0. Multivariate modeling fitting was performed using the ‘lme4’ package [ 55 ] in the R Programming Language 4.1.0 [ 56 ]. Initial univariate comparisons of the forensic anthropology caseloads from the three agencies (UNT, FBI, NYC) were completed, including a comparison of the NamUs region of origin and the decedent demographic data distributions for each agency (sex, age, and race and/or ethnicity). Chi-square tests for agency distribution differences were carried out to compare the three samples. Identification status (identified or not identified) trends were then generated and compared across decedent demographic variables, and Chi-square tests (or Fisher’s Exact test when appropriate) and correspondence analyses were completed to compare the demographic distributions of identified versus unidentified individuals. We also examined the availability of biological profile information as it relates to identification rates.

Wilcoxon Rank Sum/Kruskal Wallis tests were used to infer whether there were significant differences in investigation duration (length of time to identification in years) across decedent demographic groups. This analysis requires an identified case sample and was completed using the UNT sample because both recovery date (by year) and identification date (by year) were only available from this agency. For the UNT identified case sample, the “investigation duration” was determined to be the total number of years from case recovery to identification.

Importantly, the analyses of investigation duration would inevitably include cold cases from many decades ago, and the prevalence of such cases will be dependent on the trends in U.S. demographics over the years, which could bias the sample if left unchecked and produce substantially different mean investigation durations. For example, while there are substantial cases in our dataset representing White decedents from 20 to 50 years ago, the majority of cases representing Hispanic decedents occur in more recent decades. As another example, improvements to investigations in more recent decades could produce shorter investigative durations. In order to mitigate the effects of these factors on identification duration, we took the following steps. For the identified case analysis of the relationship between investigation duration and decedent race, only Hispanic and White had sufficient sample sizes for comparison. Based on the case recovery year distributions, we found that Hispanic decedent cases became more prevalent in the UNT sample more recently, especially over the past decades, with the 90% (10% quantile) of all UNT Hispanic cases occurring in or after 2012, and the 75% (25% quantile) occurring in or after 2014. These were therefore used as the case year cutoffs in order to ensure comparable year distributions and prevalence for the White and Hispanic decedents. Furthermore, this time period allows for comparable identification modalities, as all cases are included in the period when UNT had DNA analyses regularly integrated into casework.

Mixed effects logistic regressions were employed to examine i) combinations of sex, age, and race and/or ethnicity as they relate to identification status, and ii) whether having estimates of biological profile components influences identification status. A mixed effects model taking into account the non-independence in the dataset was employed. In this case, we assume there are agency-specific identification trends (with regards to demographic factors) in addition to general trends in identification. That is, cases from one agency were handled by a particular group of practitioners with agency-specific training and experience and may be subject to the same influences. It is therefore necessary to explore agency-specific trends using a multivariate approach. Identification status was modeled as a binary outcome variable (identified vs. unidentified) while sex, age, race and/or ethnicity, and whether these components were estimated were modeled as categorical predictor variables. To ensure a fully represented data matrix, the combined demographic categories with zero cases were removed prior to model fitting. Since doing so removed all Juvenile individuals and all Native American individuals, we report both univariate and multivariate analyses results as they complement one another and provide a more comprehensive picture of the identification trends. The resulting model was then checked for collinearity between predictors as well as influential cases to ensure model assumptions are met [ 57 ]. To minimize abstraction, we describe our models in terms of comparison with the reference categories for the predictors. The reference categories are females, adolescents, and White.

Because multiple tests were performed on the same subsamples of the dataset, we used the Benjamini-Hochberg false discovery rate p- value adjustment for multiple tests [ 58 ]. To measure the effect size of the Chi-square tests, Cramer’s V was employed, using the 95% confidence interval for its estimated value. Cramer’s V is an extension of the correlation coefficient used for 2 x 2 contingency tables and ranges from zero to one, with zero implying no correlation (i.e., negligible effects) and values closer to one implying a strong correlation (i.e., large effects). Cramer’s V values can be interpreted using Cohen’s d degrees of freedom-informed approach [ 59 ]. While effect size is not traditionally used in forensic anthropology research, it has been considered a statistical best practice in many fields, including psychology, medicine, ecology and evolutionary biology [ 60 ]. However, the thresholds for interpreting what size effect is considered important is known to be relatively arbitrary unless there are other studies to compare it to. When comparative effect sizes are unavailable in the discipline, Cohen [ 59 ] proposed general thresholds for use: negligible effects ( V ≤ 0.10), small effects (0.10 < V < 0.30), medium effects (0.30 < V < 0.50), and large effects ( V ≥ 0.50). For this study, the thresholds are defined using the ranges provided above. We report effect size for three reasons. First, given the large sample sizes used in this study, the odds of statistically significant tests at p = 0.05 are extremely high, but not meaningful beyond that a relationship exists (slim chance of a Type I error), but the magnitude of the relationship between the variables of interest cannot be directly inferred from the p- value [ 61 , 62 ]. Thus secondly, reporting effect sizes for each of our analyses in the present study allows us to compare the strength of the relationships for each test and across agencies. Third, the effect sizes reported here can be used as the first comparison for future studies using effect sizes in our discipline when studying similar topics.

Comparing agency datasets: Demographics, geographic distributions, and identification rates

The three agencies have substantial differences in their regional representation ( Table 2 ), with the FBI sample being the most representative of the nine NamUs regions. The UNT sample is primarily represented by Region 4 (73% of its forensic anthropology caseload), but does have cases from all but Regions 8 and 9. In contrast, 100% of the NYC sample is from Region 6, the majority from New York City. When all three agency samples are combined, all NamUs regions are represented, although not evenly (e.g., Regions 4 and 6 comprise 54% of the total sample).

Region 1 (AZ, CO, HI, NM, NV, UT, Guam, Saipan), Region 2 (AK, ID, MN, MT, ND, OR, SD, WA, WY), Region 3 (IA, IL, IN, KS, MO, NE, OH, WI), Region 4 (AR, LA, OK, TX), Region 5 (AL, KY, MS, PA, TN, WV), Region 6 (CT, MA, ME, MI, NH, NY, RI, VT), Region 7 (FL, GA, NC, SC, Puerto Rico, US Virgin Islands), Region 8 (DE, MD, NJ, VA, District of Columbia), Region 9 (CA).

https://doi.org/10.1371/journal.pone.0290302.t002

Comparisons of the distributions for decedent sex, age, and race and/or ethnicity are presented in Table 3 . Regarding decedent sex, all agencies had at least twice the frequency of male decedents compared to female decedents. The FBI and NYC samples had comparable proportions, while UNT presented a slightly greater proportion of male decedents compared with the other two agencies. Age trends indicated that the UNT sample more frequently included young adults and less frequently included older adults compared to the NYC and FBI samples. When considering the combined frequencies of the three agency samples, juvenile and adolescent cases were the least common, while young adults and middle-aged adults were comparable in frequency. The greatest frequency differences in decedent demographic trends among the three agencies were for decedent race and/or ethnicity. The majority of UNT’s sample was comprised of Hispanic (55%) decedents, followed by White (17%) decedents. In contrast, the majority of NYC’s forensic anthropology sample was comprised of White (35%) and Black (25%) decedents. The majority of FBI’s sample was comprised of White individuals (43%), which were about two to three times as common as any other decedent race and/or ethnicity. For all agencies, Native American decedents and Asian decedents were the least frequent (0 to 5%).

https://doi.org/10.1371/journal.pone.0290302.t003

Chi-square tests of agency-specific differences for the frequency distributions of decedent sex, age, and race and/or ethnicity were performed, excluding those cases with an inconclusive decedent sex, age, or race and/or ethnicity. Furthermore, the Native American subsample was excluded from the Chi-square test for race and/or ethnicity because the sample sizes were not substantial enough and left some cells with zero cases. All tests for agency differences in decedent sex, age, and race and/or ethnicity frequencies yielded statistically significantly results ( Table 4 ). The accompanying Cramer’s V statistics indicated that decedent sex and age variation have a limited relationship (i.e., small effect) with agency of origin, while race and/or ethnicity variation appears to be more strongly related to agency of origin, with a medium effect.

Shaded cells indicate statistical significance of test for the adjusted p values.

https://doi.org/10.1371/journal.pone.0290302.t004

We compared the identification rates among the three agency samples ( Table 3 ) and found that identification rates ranged from 41% to 57%, with NYC identification rates being the greatest. The Chi-Square test for frequency differences in identification rates among the three agencies was a significant ( Table 4 ), but small effect. When pooled, the identification rate for the sample was 47%.

Aim 1: For forensic anthropology cases, do investigation trends (e.g. identification success rate; duration of an investigation) differ across demographic backgrounds

Pooled sample univariate analyses..

Using the pooled sample, we then explored the relationship between identification status (either identified or not identified at the time this data was collected) and its associated decedent demographic data (sex, age, and race and/or ethnicity). Cases where inferred decedent data for sex, age, or race and/or ethnicity were lacking were excluded from each analysis. The pooled samples’ Chi-square results for examining identification rate trends across decedent demographics are presented in Table 5 , and the identification frequencies associated with demographic data are presented in Fig 1 . Chi-Square tests and Cramer’s V were employed to identify any statistically significant and notable trends. The pooled sample’s results suggest that identification frequencies statistically differed with decedent age and sex, but not decedent race and/or ethnicity. The Cramer’s V measure of effects for age and sex effects on identification status are small and negligible, respectively, thus indicating limited differences among the age and sex subsamples’ identification rates. For sex, females (54% identification rate) were more frequently identified than males (47% identification rate). Fig 1 also highlights that the main frequency differences for age indicate that juvenile, adolescent, and older adult subsamples have higher rates of identification when compared with middle and young adult subsamples.

https://doi.org/10.1371/journal.pone.0290302.g001

Pooled and agency-specific tests performed. Shaded cells indicate significance of adjusted p -value.

https://doi.org/10.1371/journal.pone.0290302.t005

Agency-specific analyses.

Because pooling the sample could be masking agency specific trends, we again examined the relationship between identification rates and decedent demographic data (sex, age, and race and/or ethnicity) per agency sample. The identification rates per agency are reported in Fig 2 , and the Chi-square results are presented in Table 5 . We found that none of the three agencies yielded significantly different identification rates for males and females. For decedent age, we found that only the FBI sample’s identification rates significantly differed among age subsamples. While there were no significant differences in identification rates among age subsamples for UNT and NYC, their trends indicate agreement among the agencies in regard to identification rates, with Older Adult subsample having the greatest identification rates (74% and 72%, respectively). For the FBI and NYC samples, the Adolescent case subsamples had some of the highest identification rates (64% and 69%, respectively). The Middle Adult subsample exhibited the lowest identification rate for the FBI (47%) and NYC (51%) samples, and the second lowest for the UNT (49%) sample.

https://doi.org/10.1371/journal.pone.0290302.g002

When examining the relationship of decedent race and/or ethnicity with case identification rates, agency-specific datasets did not adequately represent all available subsamples used in the pooled sample. For NYC, Native American decedents were excluded, and for UNT Native American and Asian decedents were excluded. The FBI dataset had no groups excluded. The only agency with statistically significant differences in identification rates across race and/or ethnicity was the UNT sample ( Table 5 ), with a medium effect size. For the three groups examined (Black, Hispanic, and White), White decedents’ identification rate (81%) was twice that of Hispanic decedents’ (42%), while the Black decedent identification rate was intermediate (67%). In contrast, NYC Hispanic decedents present the greatest identification rate (81%) compared to Asian (61%), Black (69%) and White (66%) decedent identification rates. The FBI analysis had highly comparable identification rates for Black, Hispanic and White decedents (47–51%), while Native American and Asian identification rates were the greatest (63%) and lowest (21%), respectively. Given that the only agency with significantly different identification rates among race and ethnicities is the UNT sample, this suggests that the statistical significance of this same test when using the pooled sample is likely driven by the differences found within UNT, and not a nation-wide trend of lower identification rates for Hispanic decedents. It should be noted that the small samples for decedents—namely, FBI Asian (n = 14), FBI Native American (n = 16), and UNT Black (n = 15)—may not be representative of United States identification trends at large for these subsamples, and more data will be needed to better gauge the identification trends.

Pooled multivariate analyses.

While the univariate tests indicate trends for the independent factors of decedent sex, age, and race and/or ethnicity, it is also important to ensure that we take into account the multivariate nature of demographic data, as individuals simultaneously belonging to a specific combination of sex, age, and race and/or ethnicity for any given case, as well as considering agency of origin. This approach most strongly corresponds with the reality of casework, in that a decedent’s sex, age, and race and/or ethnicity would typically be simultaneously considered in a case and investigated within a particular agency. Thus, we employ a multivariate approach to more realistically infer the relationship between identification rates and decedent demographics. The multiple logistic regression model indicated the overall effects of decedent sex, age, and race and/or ethnicity were not significant, after agency effects were taken into account ( Table 6 ). The results of the logistic regression model also revealed the direction of identification rate differences while taking all demographic factors into account, and the results were consistent with the univariate findings. All other factors equal, females are more likely to be identified than males, Middle Adults are less likely to be identified than other age groups, and White and Asian decedents are more likely to be identified than other race and/or ethnicity. However, the amount of overlap in conditional probabilities among different categories for each individual predictor variable (sex, age, and race and/or ethnicity) again suggest the differences in identification rates among the subsamples are almost nonexistent for sex and race and/or ethnicity, and small for age, primarily driven by the Middle Adult group’s lower identification rates deviating from all other age groups. The model also allows us to explore the combination of factors such as sex and race and/or ethnicity as they relate to identification status ( Fig 3 ). Regardless of age, White females and Asian females have slightly higher probabilities of identification compared to other subsamples. However, the amount of overlap among subsamples is big and suggests that the effects of demographic factors to identification status, if any, are small.

Note the overlaps among subsamples suggest the effects of demographic factors on identification probabilities are small.

https://doi.org/10.1371/journal.pone.0290302.g003

The reference categories (odds ratio = 1) are female, adolescent, and White.

https://doi.org/10.1371/journal.pone.0290302.t006

Investigative duration trends.

To explore whether duration of an investigation differed across decedent demographics, we examined the relationship between investigation duration and decedent sex, age, or race and/or ethnicity using the UNT identified case sample ( Table 7 ). Employing Wilcoxon Rank Sum/Kruskal Wallis tests, we then tested for differences in the investigation duration distribution across decedent sex, age, and race and/or ethnicity. Subsamples analyzed for these analyses were limited to those with n ≥ 20, resulting in the following groups per decedent demographic component for the identified UNT case sample: decedent sex (Male and Female), decedent age (Young Adult, Middle Adult, Older Adult), and decedent race and/or ethnicity (Hispanic and White). Ninety-eight percent of the cases used in this analysis were DNA identifications, meaning that when holding identification modality constant, we can explore the effects of demographics on investigation duration. It has been hypothesized [ 1 , 34 ] that anthropological estimations of sex, age, or ancestry could influence investigations such that some demographics’ cases are prioritized over others resulting in lengthier investigation times for deprioritized groups. Therefore, in Table 7 we also report the percent of cases with corresponding biological profile estimation information available in the UNT samples analyzed for investigation duration trends. The vast majority of cases had their biological profile information available at some point in the investigation, and therefore it can be assumed that if this information was to bias the investigation in some way, we would see that emerge as differences in the investigation duration trends. The median investigative duration for all seven groups was one year, and the distributions of the investigative durations did not significantly differ among the tested sexes nor among the tested age groups. Regarding the race and ethnicity, recall from the Methods section that for the comparisons of Hispanic and White decedent investigation durations we employed for two case year cutoffs, 2012 and 2014. The 2012 analysis resulted in significantly different investigation duration distributions (but identical medians at 1 year), primarily in the upper quantiles (75%-100%) of the distribution, with Hispanic decedents having greater investigative durations than White decedents (2–5 years versus 1–3 years). In contrast, the 2014 analysis did not find significant differences in investigation durations for White and Hispanic decedents.

Shaded cells indicate significance of adjusted p -value.

https://doi.org/10.1371/journal.pone.0290302.t007

Aim 2. Is the decedent information included in forensic anthropology reports (e.g., estimations of decedent ancestry, sex, and age) related to the (de)prioritization of particular individuals over others in regard to identification efforts

Decedent demographic trends are analyzed similar to Aim 1, but with the dataset comprised of only forensic anthropology estimated demographic data. In this way, we are able to compare how forensic anthropologists’ estimated demographic information about a decedent are related to investigative trends. The identification trends for only estimated decedent sex, age, and ancestry were generally comparable to those found in Aim 1 using known and estimated decedent data, with a few exceptions highlighting greater demographic discrepancies. In particular, for the pooled agencies analyses, all three demographic factors (estimated sex, age, and ancestry) yielded significant differences in identification success rates (S1 Table in S1 File ), although the Cramer’s V effect sizes were negligible for estimated sex and small for estimated age and ancestry. Decedents estimated as White and Black had the greatest investigative success with 54% and 53% of cases identified, respectively. Clustered with lower investigation success rates (39%-41%) are decedents estimated to be Asian, Hispanic, or multiple labels. Agency-specific analyses of estimated demographic data and identification rates indicated no significant relationships for the FBI sample. For NYC, there was a significant relationship (small effect) between identification rate and estimated ancestry such that decedents identified as Multiple labels were less frequently identified. For UNT, both estimated age (small effect) and ancestry (medium effect) were significantly related to identification rate, with Young Adult decedents as well as Hispanic decedents less frequently identified than other groups. Agency-specific investigation duration trend analyses indicated no significant differences in time to identification for the included UNT sample (S2 Table in S1 File ).

The multivariate results, which allow for the composite demographic estimates to be observed are somewhat different from the results of univariate analyses, showing the overall effects of estimated sex, age, and ancestry were not significant, after agency effects were taken into account (S3 Table in S1 File ). This is consistent with the multivariate results for Aim 1. For the combination of estimated sex and ancestry and how they relate to identification status (S1 Fig in S1 File ), the 95% confidence intervals for the probabilities of identification all overlap. The wide identification probability CIs associated with decedents estimated to be Asian females and females with multiple ancestry labels are likely associated with their small sample sizes, and thus the lower probabilities of identification need further investigation with more robust samples. For the other six group, White and Black males and females have substantial overlap in their 95% CIs for identification probability, while Hispanic males appear to be the lowest and have limited overlap with the White males. This trend is consistent with odds ratio analyses (S3 Table in S1 File ), where the only statistically significant odd ratios was White decedents being twice as likely to be identified as Hispanic decedents.

Aim 3. Does providing information related to the decedent’s age, sex, and race impact the odds of an identification?

We examined whether biological profile estimates provided by the forensic anthropologist (e.g., sex, age, or ancestry) have a relationship with identification rates. We performed a mixed effects logistic regression to infer whether identification rates related to whether or not forensic anthropology decedent demographic estimates were provided (e.g. “provided” or “not provided”), taking case agency into account. The odds ratios for the presence of biological profile components are as follows: 2.16 (95% confidence interval: 0.72–6.44, p > 0.05) for sex, 1.21 (0.47–3.15, p > 0.05) for age, and 1.05 (0.47–2.34, p > 0.05) for ancestry) indicated that estimations of each biological profile component only had a limited relationship with the probability of identification. When all three components were estimated (compared to when no component was estimated), the probability of identification increased twofold ( Fig 4 ).

An increase in identification probability is evident when all biological profile components are estimated (26.7% probability of identification with no biological profile component estimated vs. 50.1% probability of identification with all biological profile components estimated).

https://doi.org/10.1371/journal.pone.0290302.g004

To better understand which agencies are contributing to the observed differences in identification rates in the multivariate models, we examined the relationship between identification status and the presence or absence of forensic anthropology estimates of sex, age or ancestry for the agency-specific samples using univariate tests ( Table 8 ). We found that NYC identification rates significantly improved when sex, age, or ancestry estimations were provided by 40%, 27% and 25%, respectively, compared to when these estimations were not provided, but all were considered small effects. In contrast, the FBI and UNT presented no significant differences in identification rates related to the presence or absence of sex, age, or ancestry estimations.

The "n/a" status for the UNT age test is because age was estimated in all cases and thus the test could not be performed. Shaded cells indicate statistical significance of test for the adjusted p values.

https://doi.org/10.1371/journal.pone.0290302.t008

We performed a follow-up analysis which explored whether having an ancestry estimation available (two levels: ancestry estimation provided; ancestry estimation not provided) differentially impacted the identification outcome across decedent race and/or ethnicity. This analysis was performed to explore whether trends exist that suggest cases are (de)prioritized by the ancestry information that the forensic anthropologist provides. For example, if unidentified persons investigations are biased towards prioritizing cases of White decedents, then we would expect to see a greater proportion of White decedents in the sample of positively identified cases with ancestry estimations, compared to the positively identified case sample without ancestry estimations. Therefore, we analyzed whether the proportion of race and/or ethnicity groups differed between identified individuals with ancestry estimations versus identified individuals without ancestry estimations. In this scenario, the subsample of identified case without an ancestry estimation is considered the control sample, as there is no information on the decedent’s possible race and/or ethnicity to potentially bias the investigation, and thus the proportions of race and/or ethnicities in this subsample will be considered the control. Alternatively, the identified case sample with ancestry estimations can be considered the biased case subsample, as there is information on the decedent’s possible race and/or ethnicity, which has the potential to bias the investigation.

For this analysis, identified individuals representing three decedent races and/or ethnicities (those with substantial sample sizes being Black, White, and Hispanic decedents) from all agencies were used. We found that race and/or ethnicity proportions of identified cases ( Fig 5 ) were not significantly related to whether or not an ancestry estimation was reported ( N = 522, adjusted p = 0.066; Cramer’s V = 0.106). While not significant, it is interesting to note the trends observed, with both Black and Hispanic decedents exhibiting higher proportions (9% and 6% higher, respectively) in the identified case subsample when ancestry was estimated, while White decedents had a concomitant decrease (14% decrease). Similar to the trends of the pooled sample, the agency specific analyses showed an increase in the identification of Black decedents when ancestry is reported. However, only UNT exhibits significant differences between Black, Hispanic and White decedent proportions for identified individuals with versus without an ancestry estimation (Fisher’s Exact Test, N = 143, p = 0.028 Cramer’s V = 0.225). For UNT, the identification of Hispanic individuals almost doubles (to 63%) when ancestry information is reported, compared with White decedents whose frequency is almost halved (30%).

https://doi.org/10.1371/journal.pone.0290302.g005

Similar analyses could not be performed for decedent sex or age because there are only a limited number of identified decedents without forensic anthropology estimates of age or sex, and thus inadequate comparative data of identified decedents with or without this biological profile information. An unreported biological profile estimation may be due to several factors, including but not limited to incomplete and/or damaged skeletal material, methodological outcomes with weak probabilities, or contradicting outcomes from various methods.

This study examines the broader context of identification trends for forensic anthropology casework, as it relates both to decedent demographics and the information provided by the forensic anthropologist (the biological profile estimations). We examined these trends for both pooled and agency-specific samples because agency contexts (e.g., jurisdiction of cases, decedent demographics of casework) significantly differ among the three datasets included in the present study.

Do identification trends vary across decedent demographics?

In Aim 1, we examined the relationship between identification rates and decedent demographics (known and estimated) using both univariate and multivariate approaches. The multivariate logistic regression offered the most comprehensive analysis in that it allowed for the simultaneous consideration of sex, age, and race and/or ethnicity of each decedent as well as the agency effects. The model indicated the overall effects of decedent sex, age, and race and/or ethnicity on identification rate were not significant ( Table 6 ). These findings corroborate the univariate analyses of the pooled sample, which indicated that while both decedent age and sex significantly related to identification rate, their overall effects on identification rate were small or negligible, and appear to become insignificant when considered in tandem with other relevant factors in the multiple logistic regression. When comparing to the agency-specific univariate tests where the only significant finding was a relationship between identification rate and decedent race and/or ethnicity for UNT, this factor did not remain significant in the multiple logistic regression when agency differences were taken into account. These Aim 1 analyses suggest that decedent sex, age, and race and/or ethnicity have a limited (univariate analyses) to negligible (multivariate analysis), relationship with identification rates for the study samples.

Beyond identification rate, duration of the investigation for resolved cases was examined in Aim 1. The significant increase in investigation duration for UNT cases from 2012 to present for Hispanic decedents compared to White decedents compared to the lack of significantly different investigation durations for Hispanic and White decedents by 2014 may be explained by the changes of undocumented migration patterns around this time. Deaths of undocumented migrants in Texas were substantially increasing around this time, surpassing Arizona’s numbers in 2012 [ 63 ]. The influx of undocumented migrant deaths could have impacted UNT’s caseload, with needed investigative adjustments to accommodate the rise in this decedent demographic. By 2014 (through 2021), we found that the investigation durations were no longer elevated for Hispanic decedents (compared to White decedents), which could correspond with the sustained uptake of investigative strategies by the agencies served by UNT and the UNT Forensic Anthropology Unit itself. Interestingly, even though UNT’s Hispanic and White decedents exhibited the most significant difference in identification rates , in more recent years (2014 to present) when investigations do yield an identification, Hispanic and White decedents do not appear to significantly differ in the amount of time it takes to make those identifications. Furthermore, comparable investigation durations across subsamples cannot be explained by cases that come in with tentative identifications nor the modality of the identification, as those were controlled for. For future studies, examining this variable using a finer scale, such as months instead of years, and with a larger data set from more agencies would be ideal to corroborate the trends found with the available UNT data.

Do identification trends vary across estimated decedent demographics?

Aim 2 performed similar analyses to Aim 1, but with a sample whose decedent information was entirely based on the forensic anthropology biological profile estimates. In this way, Aim 2 explicitly examined the relationship of identification trends with the estimated decedent demographics data, which are what investigators would have been privy to during their investigations. Recall that it is this estimated demographic information that Bethard and DiGangi [ 1 ] proposed could influence the way in which an investigator (de)prioritizes one case over another. In general Aim 2’s univariate trends were similar to the Aim 1, with small to negligible effects for most estimated decedent demographics, and with a medium effect for estimated ancestry. Aim 2’s multivariate analysis of identification rates across various estimated decedent demographics indicated that only ancestry was influential, but the only statistically significant odds ratio for being identified (S3 Table, S1 Fig in S1 File ) was for decedents estimated as Hispanic versus White at 0.44. This finding is consistent with the univariate results for Aims 1 and 2 where White decedents were identified at twice the rate as Hispanic decedents for UNT cases. While not significant, the FBI’s analysis also indicated that Hispanic decedents had the lowest rate of identification among those analyzed. However, agency variation is important to acknowledge, as Hispanic decedents had the highest identification rates in NYC at 75%, which was the greatest identification rate for any decedent ancestry, race, and/ethnicity examined. Furthermore, similar to Aim 1, even within an agency with a low rate of identification of Hispanic cases like UNT, Aim 2’s analysis of investigation duration did not result in statistically significant differences for the three compared groups (S2 Table in S1 File ). Collectively, Aim 2 exhibited more statistically significant differences in identification rates than Aim 1, suggesting that the particular information provided by forensic anthropologists has a relationship with identification outcomes.

Does the availability of biological profile estimations relate to identification trends in forensic anthropology cases?

We also addressed the question of identification discrepancies when biological profile estimates are provided to investigators. Through univariate and multivariate testing, we found that identification rates appear to have a significant but limited relationship with the availability (e.g. presence or absence of this information) of biological profile estimations. Trends from the pooled sample’s multiple logistic regression indicate that decedents with all three biological profile components present in a case report have a greater probability (50%) of being identified compared to decedents who have none of these biological profile components estimated (27%). Agency-specific univariate analyses yielded further insight ( Table 8 ), highlighting that NYC was the only agency where the presence of sex, age, and ancestry estimations each independently significantly improved identification rates by 40%, 27% and 25%, respectively.

The relationship between forensic anthropological analyses/involvement with a case and identification rates can also be considered within the nationwide context of identification rates. The pooled FBI, NYC, and UNT identification rate (47%) is almost double the rate of identifications of unknown individuals uploaded to NamUs. Out of the 18,740 unidentified individuals uploaded, approximately 27% have since been identified (data accessed on 7/28/2021). Interestingly, cases of unidentified individuals uploaded to NamUs also include individuals with a tentative name upon the initiation of the investigation which presumably have a better chance of being identified than those without a lead. However, individuals with a tentative name were excluded from the present study, which indicates that the identification rates reported here are likely an underrepresentation of the overall identification rates for U.S. forensic anthropology cases.

One key difference between NamUs and the present study is that NamUs includes all cases of unidentified individuals, while the present study only includes unidentified individuals with a forensic anthropological biological profile (though not necessarily complete). Any discrepancy in overall identification rates between NamUs and the present study could also be tangentially related to the agencies’ represented in the two datasets—specifically, that agencies that hire in-house forensic anthropologists are inherently larger with more local resources (e.g., NYC cases and some of the FBI and UNT cases) to put towards investigations in general, which may indicate that these agencies have a better chance of identifying individuals. Additionally, those agencies using external forensic anthropologists (such as the agencies represented in the FBI and the UNT samples) could also represent smaller but actively engaged agencies with a better understanding of how to leverage the available resources at the local and national level (e.g., UNT and some of the FBI cases), which may in turn increase their odds of identification. Finally, the discrepancy in identification rates between NamUs and the present study could be directly related to the role of the forensic anthropologist, in that the estimation of a biological profile may increase the chance of identification and/or the kind of general engagement forensic anthropologists have with casework improves the odds of an identification. Because there are potentially confounding factors in an investigation that could result in a spurious association of the availability of forensic anthropology estimates and the chance of identifications (e.g., poorly preserved/recovered skeletal remains), we cannot infer a causal relationship from these results. While the effects size for the NYC univariate tests were small, the significant results of the multivariate analysis (which took agency differences into account) and the consistency with broader national trends suggest improvements to identifications when the biological profile is provided and that more research would be useful to infer whether this positive impact (though small effect) of forensic anthropology reporting holds beyond the present study sample.

Lastly, we addressed the specific question of identification discrepancies when ancestry estimates are provided to investigators by examining whether identification trends vary with the availability of reported biological profile estimates. After the multivariate finding that the availability of biological profile information has a significant relationship with the odds of an identification, it is important to understand whether biological profile information has the same relationship with identifications for different decedent demographic groups. Namely, we examined the postulation that biological profile estimations of ancestry can inadvertently prioritize identifications of one decedent over another, specifically cases of White decedents [ 1 , 34 ]. In this context, any case where ancestry was estimated has the potential to bias investigators in the direction of prioritizing White individuals and deprioritizing all others, which would result in a larger proportion of White individuals in the identified case subsample with ancestry estimations, and a smaller proportion of White individuals in the identified case subsample without ancestry estimations. Any identified individual without an ancestry estimation would have had no information that could bias the investigation toward prioritizing decedents of a certain race over others. Using the identified case sample, we found no differences in decedent race and/or ethnicity proportions when ancestry was and was not estimated ( Fig 5 ), suggesting limited evidence that prioritization of certain groups over others is occurring for the present study sample for this analysis.

Interestingly, in the agency-specific analyses, UNT exhibited a significant trend that is opposite of the exhibited identification rate trends of White decedents twice as frequently identified as Hispanic decedents. When ancestry was estimated, the proportion of Black and Hispanic individuals increased , while the proportion of White decedents decreased when compared to unidentified remains cases where ancestry was not estimated. This trend suggests a clear relationship between decedent race and/or ethnicity, identification, and available ancestry estimations. While prioritization of Black and Hispanic individuals in investigations is one way to interpret these findings, an alternative explanation is that estimations of ancestry improve the odds of an identification for Black and Hispanics decedents more so than for White decedents. This aligns with previous work [ 64 ] that indicates the evidentiary value of an ancestry estimation towards an identification is largely based on the racial diversity of the pool of missing persons. Given that UNT cases originate from all over the United States, the nationwide pool of missing persons is largely represented by White individuals (59%), making an ancestry estimation related to White the least unique/lowest evidentiary value for contributing to an identification. Alternatively, ancestry estimations related to Black and Hispanic decedents would have greater evidentiary value, as Black and Hispanic individuals comprise only 16% and 11% of the missing person pool, respectively (NamUs [ 65 ]). However, if evidentiary value was the main factor contributing to the observed differences, then we should have found a similar trend in the FBI as we did with UNT, given that the FBI sample also consists of nation-wide cases. Other factors, such as the lack of missing persons report being filed would also impede the ability for biological profile estimations to assist in an identification. Further study of these trends and potential factors contributing to them will help to elucidate the relationship among decedent race and/or ethnicity, the availability of ancestry estimations, and identifications in casework.

Providing context: Factors influencing investigative trends

In this study, the majority of investigative trends related to decedent demographics are negligible to small effects. Notably, decedent ancestry, race, and/or ethnicity exhibited medium effects for Aims 1 and 2 in the univariate analysis for the UNT sample. These findings provide evidence consistent with our hypotheses for Aims 1 and 2, in that decedent demographics are related to identification rates. One interpretation of these findings is to conclude that (de)prioritization of cases based on the decedent demographic is what is driving these results. However, the results were inconsistent across univariate and multivariate tests, as well as at times contradicting when comparing various measures of (de)prioritization we employed. For example, while identification rates appear to be significantly related to decedent demographics, identification duration did not always demonstrate similar results. Furthermore, Aim 3 found no differences in decedent race and/or ethnicity proportions subsamples when ancestry was and was not estimated ( Fig 5 ), suggesting limited evidence for prioritization. Collectively the inconsistency in the results are unsurprising, given that whether or not active (de)prioritization is occurring by investigators, that is just one of myriad factors which contribute to each cases’ probability of being identified. In the following paragraph, we discuss some of these factors and how they offer additional considerations beyond only (de)prioritization to contribute to the observed trends in our study.

When considering decedent sex, females exhibited significantly greater identification rates compared to males, although the overall difference is less than 10% with a negligible effect per Cramer’s V . There are several contextual factors that could play a role in the increased identifications of females, such as females being reported missing more often than males, and/or overall fewer missing and unidentified females such that there is inherently a smaller pool of potential matches in the investigation of unidentified females. National statistics on missing and unidentified persons reflect these trends of greater reporting yet lower prevalence among the unidentified, with females only comprising 19% of the unidentified persons in NamUs, yet 39% of the reported missing persons (NamUs [ 65 ]).

Age groups were found to be significantly related to identification rates for the pooled sample, with Juveniles, Adolescents and Older Adults having better identification rates than Middle and Young Adults. However, this same trend was not found to be significant in the multiple logistic regression, nor the agency-specific univariate tests, with the exception of the FBI sample which showed only a small effect driven by the Adolescent group outperforming (greater identification rates) Young, Middle, and Older adult groups. It is possible that law enforcement may not always take missing persons reports for missing young and middle-aged adults because they could be viewed as going missing voluntarily, thus contributing to these cases remaining unidentified. In addition, juvenile, adolescent and older adult age subsamples are intentionally included in various societal systems of surveillance that work simultaneously to aid in identifications and may in part explain the improved identification rates compared to other age subsamples. According to a 2021 National Center for Missing and Exploited Children (NCMEC) report [ 66 ] in which identifications of deceased children were analyzed, individuals aged 11–20 represent the largest percentage of deceased and identified children (65%)in the study. Thus, among child/adolescent deaths, this age range is both the most commonly deceased and the most commonly identified. Children in this age range have various forms of surveillance as part of their daily lives (e.g., school, parents, caregivers), and as such their disappearance is commonly noted within this system. According to the NCMEC report, the majority of deceased children were recovered in the same city or state where they went missing, and identifications primarily stemmed from law enforcement tips. This “closed system” empowers these local systems of surveillance and potentially increases identifications. In addition, cases of child/adolescent death/disappearance are subject to multiple local agencies’ investigations (law enforcement, child protective services) and may receive assistance from national entities (NCMEC, Amber alerts etc.) all working simultaneously to increase identifications.

Older adults may also have a similar system of societal surveillance related to where they die and who is looking for them. According to Muramatsu and colleague’s [ 67 ] study on place of death for older Americans, 33% died in nursing homes, <25% died at home, and the remainder died in either hospice or hospital care. Thus, the largest majority of older adults have a direct relationship to surveillance via care. Older adults often have consistent care appointments, or relationships with targeted programs like meals on wheels that surveille well-being in some metric. In addition, for older adults that may go missing, the American Silver Alert system is a public notification system countrywide that issues alerts on missing seniors, especially those that may have dementia or other cognitive impairments. Together these systems of surveillance amount to an established network of social programs that may work collaboratively to increase the identification of both adolescents and older adults in the United States.

In regard to decedent race and/or ethnicity, Aim 1 found that the pooled sample yielded no significant relationship between identification rates and decedent race and/or ethnicity s, but UNT’s agency-specific test was significant ( Table 5 ), and correspondence analysis indicated that the difference was driven by the higher identification rates for White decedents (81%) compared with Hispanic decedents (42%). UNT has more young adult, Hispanic male cases than both NYC and FBI, which is a common demographic of decedents believed to have died during an attempt to cross the U.S.-Mexico border, and multiple studies have documented the difficulties in successfully identifying decedents within this context [ 68 – 74 ].

Martinez and colleague’s work [ 70 ] examining Pima County Office of the Medical Examiner (PCOME) case data, highlighted a relationship between decedent demographics and identification rates, namely that as the number of non-Mexican, Central American individuals crossing the border increased in recent years, so too did the number of unidentified remains cases. The authors posited that poverty, consulate resource limitations, and the strategies of human smugglers along the border all increase the complexity of the investigation and create disparities in the odds of a successful identification of decedents from Central America as compared to more traditional migration regions like Mexico. Hughes et al. [ 69 ] and Algee-Hewitt et al. [ 68 ] found a similar identification disparity in PCOME forensic casework related to the ancestry of deceased individuals along the U.S.-Mexico border, in that identified migrants tended to have more European ancestry, while unidentified migrants tended to have less European ancestry. Importantly, the identification bias was unlikely to be attributed to any direct action (or inaction) by the investigators, as these particular ancestry estimates were unknown to the investigators (and the forensic anthropologists working the cases) at the time of the investigations; the ancestry estimations were instead derived from CODIS markers and/or craniometric analyses by the authors well after most of the cases had undergone their initial investigations.

These studies represent a context where all the cases were considered to be a part of a vulnerable demographic—undocumented individuals migrating across the U.S. Mexico border—yet there were still biases in the odds of being identified. Similar to Martínez et al. [ 70 ], these studies proposed that structural vulnerabilities for individuals migrating from Mexico have emerged from the economic, political, and social processes in Mexico and the United States such that certain populations (in this case, more rural, southern Mexico, and/or Indigenous) were less often able to access and/or trust government and non-governmental social services that aid in the identification of missing loved ones, such as the collection of DNA family reference samples or filing of missing persons reports, which are key to making identifications. Indeed, this finding is echoed by research focusing on how sociohistoric factors and structural vulnerabilities play a role in investigation outcomes for undocumented Latinx individuals [ 68 , 71 ]. These previous studies, along with the present study’s findings that only the UNT sample exhibited race and/or ethnicity disparities in identification rates, strongly suggest that context-specific dynamics can potentially influence the odds of an identification. As a result, the large number of U.S-Mexico border cases may inherently negatively impact the identification rates for UNT’s Hispanic subsample. However, data on the exact ratio of UNT Hispanic cases from NamUs Region 4 (which includes Texas) that are related to the border crisis was not available for the present study, and therefore we cannot conclude that these trends related to identification rates are solely a result of the United States-Mexico border context. More research on these trends may elucidate the most significant sources of the differing identification rates.

Does agency context matter?

Agency differences in identification trends highlight that the practice (and its outcomes) of forensic anthropology will inherently vary according to the context and mandates of each agency. This has important implications for research and casework, in that there may not be nation-wide trends, or, while inferable, may not be useful for understanding the ground-level practice of forensic anthropology in the United States. In this study, we have worked to incorporate models that can investigate identification trends while simultaneously accounting for agency-specific nuances. Here, we review some of the factors that may contribute to the differences we observed in identification trends across the agencies.

We found that NYC exhibited outcomes that suggest a consistent level of high performance that merited deeper examination. NYC had the greatest identification rate and exhibited a positive significant effect of reporting the biological profile on identification rate. There are substantial differences in the three agencies’ sample contexts and investigative approaches that may contribute to these trends. Both UNT and FBI forensic anthropology services span almost the entire country and they provide consultations to medical examiner/coroner and law enforcement agencies nationwide, and it is possible that the FBI and UNT may be getting a subset of cases that are inherently more difficult to identify (i.e., incomplete skeletonized remains). The FBI and UNT datasets primarily include forensic anthropology cases which are under external jurisdiction, and therefore UNT and FBI’s involvement in their cases is limited to the services requested by the originating agency. In contrast, the NYC sample in general represents cases from New York City, and within the jurisdiction of the NYC OCME. Thus, NYC OCME is in complete control of the entire investigative process, with the forensic anthropology unit deeply embedded in the investigations. These are key differences in that the results here are not truly a reflection of UNT and FBI as investigative agencies. Instead, the trends reported here for UNT and FBI reflect nation-wide trends of investigative agencies that use FBI and UNT services. Because data on the investigative agencies with jurisdiction for the cases included in the FBI and UNT samples are not available, we are unable to discuss the agency-specific practices and contexts. In contrast, because the NYC sample is by and large specific to NYC OCME jurisdiction, we focus here on the practices and investigation processes at NYC OCME in relation to their observed identification trends. We review additional factors that may contribute to identification trends across agencies in the Supporting Information.

Conclusions

The discipline of forensic anthropology has recently begun to more deeply reflect on the impact of our practices. Here, we contribute to this essential work by examining how forensic anthropology estimations of biological profile components and decedent demographics are related to identification trends. The results presented in this study indicate that for forensic anthropology cases: i) there is limited consistent evidence that decedent sex, age, and race and/or ethnicity are related to identification trends in the pooled United States sample; ii) when substantial differences (e.g. medium effects) do occur, they appear to be more agency-specific and related to the particular contexts of that agency; iii) there is no consistent evidence for discrepancies in the duration of an investigation based on a decedent’s sex or age, while race and/or ethnicity need more investigation; iv) forensic anthropological estimations of sex, age, and ancestry appear to have a small positive relationship with improved identification rates for casework.

Importantly for the present sample, forensic anthropology cases do not appear to exhibit consistent identification disparities other than those already known (such as the difficulty of identifying undocumented migrants). These results in no way negate the extensive documentation of racial bias in the criminal justice system, the possibility of delayed/deprioritization of particular investigations by law enforcement, nor the structural vulnerabilities specifically related to marginalized groups be it related to gender, race, ethnicity, socioeconomic status, or other factors. Furthermore, while our study includes cases from all NamUs regions in the U.S., our results may not necessarily determine whether disparities exist or impact identification investigations in the general U.S. practice of forensic anthropology, given that agency context and practice likely play the biggest roles in identification trends. In fact, the three study samples selected for this study reflect the best-case-scenario, in that these three agencies are state-of-the art in terms of resources, technology and quality assurance, and have multiple full-time forensic anthropologists on staff. While the best-case-scenario to practice forensic anthropology, these study samples may not be representative of all coroner and medical examiners offices in the U.S. Therefore, the current findings are a first step, and indeed, indicate that when intentional measures are taken to utilize the best investigative practices currently available, decedent demographics of deceased individuals have limited effects on reducing the chances of being identified and reconnected with their loved ones. Conversely, the findings specific to UNT’s identification rate disparities between Hispanic and White decedents suggest that some investigative contexts, such as the crisis at the U.S.-Mexico border, have hurdles that may not be fully overcome even in an agency with robust investigative means.

The findings presented here provide ample opportunity on which to expand further. Studies have hypothesized how other contextual factors that contribute to structural vulnerability and marginalization, such as gender identity, sexual preferences, race, citizenship, stigma, and socioeconomic status may contribute to identification disparities beyond the biological profile [ 5 , 20 , 69 , 75 – 78 ]. While this study found that decedent sex, age, and race/ethnicity have a limited relationship with identification trends for the study samples, other factors like socioeconomic status (SES), which are an important contributor to marginalization in the United States [ 79 ], was not taken into account. Unfortunately, there does not appear to be systematic data of the demographics and SES of individuals within the United States medicolegal death investigation system. However, based on research examining mortality in the United States more generally, SES appears to be the leading factor contributing to mortality overall, with sex and race being important factors affecting mortality within SES groups [e.g., 80 – 82 ].

Within the context of the conversation around whether and/or how to practice ancestry estimation, the results of this study offer a single data point on whether the race of a decedent is related to the odds of being identified and cannot speak to whether estimating ancestry is harming BIPOC communities within other contexts of the identification process or the lived experiences of these communities at large. Thus, the conversation is ongoing and the future discussions on ancestry estimation must still include much of what was originally called for by Bethard and DiGangi [ 1 ] as well as Stull and colleagues [ 2 ].

Using over 1,200 unidentified and identified forensic anthropology cases from three United States agencies, this study provides the first comprehensive, data-driven insight into the specific relationships between decedent data, biological profile estimations, and identification status for United States forensic anthropology cases. Even with this data, our analyses and findings are limited to the demographics that had adequate sample sizes or appropriate approaches for making inferences about the data. As such, key groups (e.g., multiracial, transgender individuals) were unable to be analyzed, and future studies will greatly benefit from a more inclusive dataset. We also acknowledge that the kind of data used in this study, namely retrospective case metrics from U.S. investigative agencies, are only one contribution to these important discussions in our discipline and is not without its own limitations. Social science approaches like ethnography will greatly contribute in ways that the data presented here cannot, and would certainly be better at measuring (de)prioritization than the proxy variables implemented here. As additional research incorporating comprehensive data continues to be developed, our understanding of biases related to the biological profile and other relevant factors will be improved, and our strategies for addressing those of significance can be refined.

Supporting information

https://doi.org/10.1371/journal.pone.0290302.s001

Acknowledgments

The authors wish to thank all three institutions (FBI, NYC OCME, UNTCHI) for allowing access to these data. The authors acknowledge the contributions of the forensic anthropologists and other medicolegal personnel employed by these institutions for their efforts in medicolegal death investigation and decedent identification. Finally, the authors acknowledge that these data represent the forensic anthropological estimates and reported demographics of deceased individuals. The use of such anonymized data allows the discipline of forensic anthropology to improve and reflect upon its’s practices.

This is publication number 22.12 of the FBI Laboratory Division. Names of commercial manufacturers are provided for identification purposes only, and inclusion does not imply endorsement of the manufacturer, or its products or services by the FBI. The views expressed are those of the author(s) and do not necessarily reflect the official policy or position of the FBI or the U.S. Government. Authors Servello, Soler, and Thomas are employed at entities from which data for this research was collected. The data that support the findings of this study are available from FBI, UNT, and NYC OCME. Restrictions may apply to the availability of these data, please contact the respective agency to inquire.

• View Article
• PubMed/NCBI
• Google Scholar
• 13. Garofalo EM, Garvin HM. The confusion between biological sex and gender and potential implications of misinterpretations. In: Klales AR, editor. Sex Estimation of the Human Skeleton: Academic Press; 2020. p. 35–52.
• 39. Goliath JR, Cosgriff-Hernandez M-TJ, editors. Increasing Disparities: Sociopolitical Implications for the Use of Ancestry Estimation in the Medicolegal System. 2021 AAFS Annaul Scientific Meeting; 2021; Virtual.
• 49. Cole LM, April K, Trinkner RJ. The Black and White Reality: Historical and Post-Ferguson Era Perspectives on Public Attitudes Toward the Police. In: Miller MK, Bornstein BH, editors. Advances in Psychology and Law: Volume 5. Cham: Springer International Publishing; 2020. p. 267–99.
• 56. R Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2021. 2021.
• 57. Osborne JW. A Practical Guide to Testing Assumptions and Cleaning Data for Logistic Regression. 2014 2021/07/30. In: Best Practices in Logistic Regression [Internet]. Thousand Oaks, CA: SAGE Publications, Inc; [84–130].
• 59. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. ed. Hillsdale, N.J: L. Erlbaum Associates; 1988.
• 65. NamUs. National Missing and Unidentified Persons System. 2023.
• 71. Reineke R, Anderson BE. Missing in the US-Mexico borderlands. In: Congram D, editor. Missing persons: Multidisciplinary perspectives on the disappeared. Toronto, Canada: Canadian Scholars’ Press; 2016. p. 246–68.
• 72. Reineke R, Halstead C. Identifying dead migrants: Examples from the United States–Mexico border. In: Laczko F, Singleton A, Black J, editors. Fatal Journeys. 3. Geneva, Switzerland: International Organization for Migration 2017. p. 77–98.
• 73. Soler A, Beatrice J, Reineke R, Martínez D, editors. Beyond identification: Structural vulnerability and the investigation of migrant deaths. 2020 AAFS Annual Scientific Meeting; 2020; Anaheim, CA.
• 78. Byrnes JF, Sandoval-Cervantes I, editors. The Marginalized in Death: A Forensic Anthropology of Intersectional Identity in the Modern Era: Lexington Books; 2022.
• 79. National Research Council. Explaining divergent levels of longevity in high-income countries. Crimmins EM, Preston SH, Cohen B, editors. Washington, D.C.: The National Academies Press; 2011.

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• PubMed/Medline
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

Recent advances in forensic anthropological methods and research.

Author Contributions

Conflicts of interest.

• Niel, M.; Chaumoître, K.; Adalian, P. Age-at-Death Estimation of Fetuses and Infants in Forensic Anthropology: A New “Coupling” Method to Detect Biases Due to Altered Growth Trajectories. Biology 2022 , 11 , 200. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Correia Dias, H.; Manco, L.; Corte Real, F.; Cunha, E. A Blood–Bone–Tooth Model for Age Prediction in Forensic Contexts. Biology 2021 , 10 , 1312. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Navega, D.; Costa, E.; Cunha, E. Adult Skeletal Age-at-Death Estimation through Deep Random Neural Networks: A New Method and Its Computational Analysis. Biology 2022 , 11 , 532. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Ross, A.H.; Williams, S.E. Ancestry Studies in Forensic Anthropology: Back on the Frontier of Racism. Biology 2021 , 10 , 602. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Butaric, L.N.; Richman, A.; Garvin, H.M. The Effects of Cranial Orientation on Forensic Frontal Sinus Identification as Assessed by Outline Analyses. Biology 2022 , 11 , 62. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Procopio, N.; Mein, C.A.; Starace, S.; Bonicelli, A.; Williams, A. Bone Diagenesis in Short Timescales: Insights from an Exploratory Proteomic Analysis. Biology 2021 , 10 , 460. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Márquez-Grant, N.; Baldini, E.; Jeynes, V.; Biehler-Gomez, L.; Aoukhiyad, L.; Passalacqua, N.V.; Giordano, G.; Di Candia, D.; Cattaneo, C. How Do Drugs Affect the Skeleton? Implications for Forensic Anthropology. Biology 2022 , 11 , 524. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• McWhirter, Z.; Karell, M.A.; Er, A.; Bozdag, M.; Ekizoglu, O.; Kranioti, E.F. Exploring the Functionality of Mesh-to-Mesh Value Comparison in Pair-Matching and Its Application to Fragmentary Remains. Biology 2021 , 10 , 1303. [ Google Scholar ] [ CrossRef ]
• Bacci, N.; Davimes, J.G.; Steyn, M.; Briers, N. Forensic Facial Comparison: Current Status, Limitations, and Future Directions. Biology 2021 , 10 , 1269. [ Google Scholar ] [ CrossRef ]
• Passalacqua, N.V.; Pilloud, M.A.; Congram, D. Forensic Anthropology as a Discipline. Biology 2021 , 10 , 691. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• De Boer, H.H.; Berger, C.E.H.; Blau, S. Providing a Forensic Expert Opinion on the “Degree of Force”: Evidentiary Considerations. Biology 2021 , 10 , 1336. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Keys, K.; Ross, A.H. Identifying Blunt Force Traumatic Injury on Thermally Altered Remains: A Pilot Study Using Sus scrofa . Biology 2022 , 11 , 87. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Garvin, H.M.; Dunn, R.; Sholts, S.B.; Litten, M.S.; Mohamed, M.; Kuttickat, N.; Skantz, N. Forensic Tools for Species Identification of Skeletal Remains: Metrics, Statistics, and OsteoID. Biology 2022 , 11 , 25. [ Google Scholar ] [ CrossRef ] [ PubMed ]

Share and Cite

Cunha, E.; Ross, A.H. Recent Advances in Forensic Anthropological Methods and Research. Biology 2022 , 11 , 908. https://doi.org/10.3390/biology11060908

Cunha E, Ross AH. Recent Advances in Forensic Anthropological Methods and Research. Biology . 2022; 11(6):908. https://doi.org/10.3390/biology11060908

Cunha, Eugénia, and Ann H. Ross. 2022. "Recent Advances in Forensic Anthropological Methods and Research" Biology 11, no. 6: 908. https://doi.org/10.3390/biology11060908

Article Metrics

Article access statistics, further information, mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals

Forensic Anthropologists | Anthropology | SIU

SIU Quick Links

• People Finder

• 618-453-2494
• [email protected]

College of Liberal Arts

Complex for forensic anthropology research, what is forensic anthropology.

Forensic Anthropology is the application of the science of biological anthropology to the legal process.  Forensic Anthropologists contribute to legal cases in a variety of ways, including recovery of skeletonized and badly decomposed human remains from the field, analysis of human remains for elements of the biological profile (age, sex, ancestry, and height), identify trauma and its potential cause, estimate the length of time the individual has been dead (the postmortem interval), and provide information that may be useful in the process of positive identification.

Complex for Forensic Anthropology Research (CFAR)

The Complex for Forensic Anthropology Research at Southern Illinois University is an outdoor research laboratory.  We focus our efforts on research, training, and service in forensic anthropology.

• The current primary research focus at CFAR is on establishing and understanding the rate and pattern of decomposition in the unique southern Illinois environment. 
• A secondary, but equally important, research focus at CFAR is on replicating common clandestine body disposal circumstances and working to understand the processes of decomposition under different disposal circumstances and how the body disposal circumstance can be identified after skeletonization.
• The growing skeletal collection allows researchers to study modern human skeletal variation, trauma, and disease.
• CFAR offers non-credit Short courses on special topics in Forensic Anthropology. Current Listings can be found here .
• CFAR offers one-day training courses to a variety of law enforcement agencies and interested professional organizations.
• The faculty of CFAR also offer traditional semester-long courses in forensic anthropology and forensic taphonomy through Anthropology.  Please see the Southern Illinois University Schedule of Classes ( https://ssb.siu.edu/prod/bwckschd.p_disp_dyn_sched ) for information on current and future term offerings.
• CFAR faculty also instruct and supervise graduate students pursuing both master’s and doctoral degrees in anthropology.
• CFAR faculty provide complete forensic anthropologic services (e.g., body recovery, skeletal analysis, trauma analysis, court testimony, etc.) to a variety of medico-legal agencies.  Please see the Forensic Anthropology Consulting page for more information about these services.
• The anatomical donation program provides an opportunity for individuals to contribute to science in a direct and meaningful way.
• Faculty and graduate students provide lectures to civic groups, school-age children, and the general public to promote science and disseminate our research results.  Please contact CFAR to arrange a lecture.

CFAR Summer Short Courses

CFAR is committed to providing high quality, professional training and education in Forensic Anthropology to students and law enforcement professionals alike, through traditional semester-long courses, short courses, and training seminars.

Through these research and educational efforts, CFAR strives to assist law enforcement with the identification of unknown deceased individuals.

Summer Short Courses

Available Summer Courses

Interested in taking a Summer Short Course?

Courses Offered :

• Introduction to Human Osteology
• Advanced Human Osteology: Introduction to Report Preparation and Courtroom Testimony

Enrollment for the summer 2019 season opens April 1, 2019

Online Enroll Form

Please direct all enrollment questions to Christiane Baigent ( [email protected] )

More From Forbes

These 6 'body farms' help forensic anthropologists learn to solve crimes.

• Share to Facebook
• Share to Twitter
• Share to Linkedin

Outdoor forensic anthropology research laboratories are colloquially called “body farms,” a term that many researchers find too sensational but that has stuck following the 1994 publication of Patricia Cornwell’s popular forensic novel “ The Body Farm .” We’re going behind the macabre moniker, though, to uncover the real story of how six facilities in the U.S. are quietly conducting pioneering research and working with law enforcement to bring killers to justice.

At these labs, scientists primarily study the process of human decomposition using donated bodies. While the general process of decomp is biologically universal, the rate of it is significantly affected by variables like temperature and humidity, not to mention by the method of disposal of the body.  If a body is found in a car trunk, a house basement, a shallow grave, or a fire pit, how can a forensic anthropologist tell time-since-death? Or discern trauma that happened before death from trauma that happened around the time of death or after death? These questions are being answered by scientists at six “body farms” around the U.S.

The first research facility of this kind was started at the University of Tennessee, Knoxville in 1981 by forensic anthropologist Bill Bass , who found the need to launch a program that studied human decomposition after being called to consult on forensic cases. After 25 years of being the only one in the country, UTK’s groundbreaking facility was followed by similar projects at Western Carolina University (2006), Texas State University (2008), Sam Houston State University (2010), Southern Illinois University, Carbondale (2012), and Colorado Mesa University (2013). These six are currently the only “body farms” that have accepted human donations, although plans are in the works for similar facilities in Florida , Wisconsin , and Pennsylvania .

We asked the directors of the six outdoor forensic anthropology research centers to share with us the most interesting and important work to come out of their centers.

Undergraduates at the University of Tennessee at Knoxville participate in a body recovery at the ... [+] Anthropological Research Facility. (Photo used with kind permission of D. Steadman.)

1. Anthropology Research Facility at the University of Tennessee, Knoxville ( Dawnie Steadman , Director).

The first of its kind in the world, the original outdoor forensic research center has grown considerably since its founding in 1981. Since it’s been around for decades, research at ARF is varied and copious, as a listing of this year’s publications shows. Law enforcement and medical examiners frequently reach out to UTK anthropologists to assist with case-based research. For example, in a case that involves a body wrapped in plastic in the trunk of a car, investigators might ask how long it takes to reach the stage of decomposition in which the body was found. This type of question can be answered scientifically through ongoing research on decomposition at the ARF.

The facility has also generated the William M. Bass Donated Skeletal Collection , which is the largest documented collection of contemporary American skeletons in the world. Researchers come from all over the globe to access it for various studies on biological profile (such as age-at-death, sex, and ancestry studies), pathologies and trauma, and occupation markers.

This collection is also part of the data set used to develop the statistical program FORDISC , which employs measurement data from skulls to help forensic anthropologists figure out sex, height, and ancestry. The most recent project that ARF is undertaking involves the investigation of mass graves, which is a major humanitarian concern in several parts of the world today. Research at the Anthropology Research Facility at UTK has revolutionized how human decomp is studied and paved the way for the development of the five other facilities below.

Cadaver dog in training at the Forensic Anthropology Research Facility at Texas State University. ... [+] (Photo used with kind permission of D. Wescott.)

2.   Forensic Osteology Research Station at Western Carolina University ( Cheryl Johnston , Director).

Researchers active at FOReSt are primarily interested in taphonomy, or what happens to a body after death, in order to gain a better understanding of how decomposition is affected by the environment and how other post-mortem processes such as scavenging affect the body itself. Students are integral to the hands-on research, undertaking projects such as figuring out whether clear or black plastic causes a body to decompose faster (black plastic does, for the curious among you) and participating in the recovery of the remains, both of which are good training opportunities. Daily photographs and videos show researchers exactly which animals contribute to decomposition, and they were surprised to find that even possums were feeding on remains.

WCU may be best known, however, for their twice yearly cadaver dog training , as they are among one of the few forensic programs in the country to offer this. Trained cadaver dogs can aid law enforcement in finding and recovering a body more quickly. Recovered remains from FOReSt are catalogued and curated at the Western Carolina Human Identification Laboratory , and both facilities hope to draw in visiting scientists living in countries where this sort of forensic decomposition research is illegal.

Student creating a facial reconstruction at a workshop at the Forensic Anthropology Research ... [+] Facility at Texas State University. (Photo used with kind permission of D. Wescott.)

3. Forensic Anthropology Research Facility at Texas State University ( Daniel Wescott , Director).

The research being done at Texas State is largely focused on estimating post-mortem interval (PMI or time-since-death) and on training law enforcement, search and rescue teams, and cadaver dogs to find remains. Both of these goals are accomplished by a focus on the interaction among species within the ecosystem where a body is found.

Vulture scavenging has therefore been a fruitful research topic, as understanding when they come, how they feed, and how to detect their presence is vital to understanding PMI. Detecting bodies is also being done here using infrared photography and drone technology, techniques that save time and reduce the number of people needed for a search. One application of this is in the search for people who have died crossing the Texas border, which FARF is contributing to along with other legal and humanitarian agencies.

Finally, the skeletonized remains that FARF collects following research are extensively catalogued and meticulously documented in terms of age-at-death, sex, health, and lifestyle to form a collection that aids researchers in both forensic anthropology and bioarchaeology.

4. Southeast Texas Applied Forensic Science Facility at Sam Houston State University ( Joan Bytheway , Director).

With both a forensic anthropologist and a forensic entomologist, STAFS’s research output involves lots of bugs, barnacles, and scavengers. Since the life-cycle of insects can be used to estimate how long a body has been dead, forensic entomology is imperative in a legal setting. Researchers at STAFS have catalogued and extensively studied the diversity of insects in the local region to find out how they might be useful in a forensic case.

Other work includes using 3D scanning to better estimate an individual’s sex from fragmentary pelvic remains, as well as studying how fire can affect human skeletal remains. STAFS is committed to educating students, law enforcement agencies, and the public about forensic anthropology. They have internships for both undergraduates and graduate students during the summer, and they offer regular short courses in forensic entomology as well as search and recovery. A recently received grant from the National Institute of Justice , in collaboration with the centers at UTK and Texas State, means more research on decomposition will be coming out soon!

A fake skull sits on a desk at the Complex for Forensic Anthropology Research at Southern Illinois ... [+] University for use as a demonstration of traumatic injury. (Photo used with kind permission of G. Dabbs.)

5. Complex for Forensic Anthropology Research at Southern Illinois University, Carbondale ( Gretchen Dabbs , Director).

Since they first accepted a human body donation in January 2012, CFAR has received 26 donations from families after death. Research has included work on vulture scavenging and the effects of freezing on decomposition, and Dabbs has discovered that bodies may naturally mummify in southern Illinois due to constant, low-speed winds.

The facility is best known, though, for a study on the effects of lawnmowers on skeletal remains. When a riding lawnmower operator accidentally ran over two research subjects, the CFAR team turned it into a new study, which they published in the Journal of Forensic Sciences .

Their latest research deals with what happens when a body is encased in concrete; namely, that the concrete may preserve the body so that it looks somewhat fresh. This finding is important because a missing persons search in this case should go back several years rather than several months. Don’t be surprised if an upcoming episode of FOX's  Bones  uses these studies for background research!

The Forensic Investigation Research Station is the facility at the highest altitude, at nearly 5,000 ... [+] feet. (Photo used with kind permission of M. Connor.)

6. Forensic Investigation Research Station at Colorado Mesa University ( Melissa Connor , Director).

This newest forensic research facility is also the one at the highest altitude -- 4780 feet above mean sea level -- and the one furthest west. Researchers have learned from some of their two dozen donations that two main things affect how fast a body becomes skeletonized: pre-death medical conditions of the donor and microenvironments in the high altitude desert. Even placement just a few yards apart can mean dramatic differences in rate of skeletonization and therefore in estimates of time-since-death. Colorado Mesa researchers work primarily with undergraduate students and also take seriously their role as “post-mortem educators,” talking with living donors who have willed their bodies to the research facility after death.

The forensic anthropologists and their students working at these six outdoor forensic research centers around the U.S. are advancing our ability to identify victims of homicide and to provide details that could help find a killer and bring that person to justice.

When he started the first “body farm” decades ago, Bill Bass never dreamed that so many people around the world would be fascinated by forensic anthropology. But the continued growth of people choosing to donate their remains to these research programs -- UTK currently has about 3,500 pre-registered donors! -- shows that people see the value of this science and the importance of this work to the community at large.

Solving murder cases and assisting in disappearances will be significantly aided by bringing forensic research centers to a greater variety of geographic and ecological locations around the country. It’s hard not to have a visceral reaction to the idea of decomposing human bodies, though. And the concerns about smell, insects, animals, and, well, gooey stuff are not unfounded. But Bass holds firm against these misgivings. “If you’re going to do things scientifically and put the bad guys away,” he says, “you have to do this kind of research.”

Kristina Killgrove is a bioarchaeologist and assistant professor of anthropology at the University of West Florida ( killgrove.org ). This post was co-authored with Tiffany Saul , a graduate student in the Department of Anthropology at the University of Tennessee, Knoxville. We appreciate the research assistance of Andy Hudson , graduate student in the Department of Anthropology at the University of West Florida .

For more information on these “body farms,” including how you can donate (your body or funding), please contact the director of the research facility nearest you.

• Editorial Standards
• Reprints & Permissions

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Forensic Sci Res
• v.6(4); 2021

Research integrity in forensic anthropology

Douglas h. ubelaker.

National Museum of Natural History,Smithsonian Institution, Washington, DC, USA

Research represents the driving force for change and advancement in the forensic sciences [ 1 ]. Of course, forensic science features casework and the application of our scientific knowledge to medicolegal issues. However, that scientific knowledge is largely based upon research conducted over decades.

Much research is stimulated by casework and the associated problems and issues that must be addressed. This is particularly true in the field of forensic anthropology in which the casework landscape has evolved rapidly in recent years. Just a few decades ago, forensic anthropologists mostly examined relatively complete skeletons in criminal cases or isolated discoveries in need of identification. Today these contexts are augmented by the analysis of mass graves in global human rights investigations, decedents in mass disasters and attacks of terrorism, and living juveniles in immigration and other legal issues. While relatively complete skeletons continue to need analysis, these are now supplemented with fragmentary evidence, bodies in an advanced state of decomposition and radiographs of the living. Anthropologists assist with conventional autopsies and in recovery operations. Demand persists for the anthropological report, but with increasing frequency anthropologists participate in more comprehensive reporting and court testimony [ 2 ].

These evolving contexts present methodological challenges for the forensic anthropologist. Anthropologists frequently find that published, accepted methods are not available to address the types of problems that must be resolved in the modern laboratory. The cases themselves and their inherent problems, each of which requires resolution, represent powerful stimuli for new research. In recent years, new research has involved thoughtful experimentation and analysis of recently acquired new collections. At times, approaches to novel methods can be found in other scientific disciplines. Three examples from my research provide a case in point.

The first relates to my over 40-year forensic career, during which time I have reported on nearly 1 000 cases, mostly at the request of the Federal Bureau of Investigation (FBI) in Washington, DC. In the early years of this consultation, cases primarily consisted of individual skeletons or major parts thereof. Toward the end of this period, cases shifted to include more fragmentary evidence resulting from fire, blast trauma and/or complex criminal activity. When small fragments were recovered and analyzed, traditional techniques were not available to distinguish bone or tooth from the many other particles that resemble these skeletal fragments and are commonly found at recovery sites. Conversations with colleagues revealed that the FBI had already developed a method using scanning electron microscopy/energy dispersive spectroscopy (SEM/EDS) to recognize different recovered materials. Through a research project, examples of bone and tooth in a variety of taphonomic conditions were added to the FBI database. The resulting new method could distinguish very small particles of bone and tooth from others and became an invaluable new analysis tool in casework [ 3 ].

The method described above allowed for recognition of small fragments of bone and tooth but it could not distinguish human fragments from those of other animals. Such recognition is extremely important in casework to allow species determination prior to any further destructive techniques (such as molecular ana­lysis). Again in the second example, extending my knowledge through conversations with colleagues and search of published literature in other scientific disciplines revealed a likely solution. Zoologists had deve­loped a method to use radioimmunoassay analysis to examine taxonomic relationships of animals. The method was developed for small fragments and was also effective on fossilized material. New research demonstrated that the method could distinguish human fragments from those of other animals and even indicate the animal species represented, if not human [ 4 ]. Once fragments were identified as bone or tooth, this new method allowed species classification, even from human fragments of great antiquity and poor preservation.

The third example relates to the estimation of time since death. Casework and research have demonstrated that many variables impact the process of postmortem change in both soft and hard human tissues. Ground water, soil acidity, presence of scavengers, constitutional factors, type and presence of clothing and containers, temperature, vegetation, and other factors all influence the timing and nature of postmortem alteration. Since many of these complex factors are unknown in forensic cases involving skeletal remains, estimation of the postmortem interval is very difficult from morphological observation alone [ 5 ]. Early in my own career, I felt confident estimating in general terms the postmortem interval from observations of tissue preservation. As more data emerged on the variables involved, however, it became clear that even general estimates were problematic.

Like many forensic anthropologists of my generation, I conducted research not only on forensic issues but also in bioarcheology. Working with archaeological samples, radiocarbon dating has provided a solid chronological framework for decades. In the past, forensic anthropologists have considered radiocarbon dating to be irrelevant for modern cases since in the few recent decades of medicolegal interest there has not been enough radiocarbon decay to measure effectively. However, when the radiocarbon laboratories report their analyses, they refer to years before present, with “present” representing the year 1950. This year reflects not only laboratory convention but also the fact that after 1950 the normally stable levels of radiocarbon in the atmosphere were drastically increased by atmospheric thermonuclear bomb testing. These augmented levels peaked around 1963 and then gra­dually decreased following cessation of atmospheric testing. It occurred to me and others that these augmented levels of atmospheric carbon-14 presented an opportunity to clarify the postmortem interval in these problematic skeletal forensic cases. If radiocarbon ana­lysis revealed the higher modern levels, then the individual must have been alive after 1950. Furthermore, research revealed that due to variance in the formation times of bones and teeth and variable turnover rates, analysis of different tissues within a single skeleton could reveal the approximate birth and death dates. Proper selection of tissues for analysis is critical and age at death is also a factor [ 6 ]. Although radiocarbon analysis has been available for many decades, we only now realize its potential to determine the postmortem interval in skeletal cases.

These three examples represent a small fraction of the robust research conducted by forensic anthropo­logists in recent years. Increasingly this vital research includes experimentation with human cadavers and/or animal surrogates, formation and analysis of documented collections of human remains and use of clini­cal data, including images and photographs of living persons. Although this research is welcomed, indeed critical to improving the practice of forensic anthropology, integrity must be maintained, and potential ethical issues must be properly addressed. The remainder of this article explores integrity and ethical concerns in this context. While this discussion focuses specifically on forensic anthropology, most issues and recommendations relate to other areas of forensic science, as well as to the diverse disciplines with whom we collaborate and learn.

Objectivity

Research, analysis, and case presentations in forensic science must remain objective. This responsibility is especially important for forensic anthropologists since their work is frequently imbedded within teams of other specialists and law enforcement personnel. The mission of forensic examination consists of objectively applying our scientific methods in the analysis and interpretation of evidence to seek the truth. This process must avoid improper influence of others who have particular theories or points of view on individual cases, whether such influence stems from political, economic or other pressure. Forensic anthropology enjoys a sound theoretical foundation [ 7 ], but safeguards must be in place to protect objectivity.

The employment situation of the anthropologist and organizational location of the laboratory can prove important in this regard. Most of my forensic research and casework has been conducted at the Smithsonian Institution in Washington, DC. Although the Smithsonian Institution operates within the federal government of the US, it offers its researchers and curators considerable freedom to conduct their work without undue influence from supervisors and administrators. This academic environment offers an ideal opportunity for independence and objectivity.

Such independence may be more problematic for anthropologists working within law enforcement, ­medicolegal teams, or other more mission-oriented organizations. In some situations, the anthropologist may need to intentionally install safeguards to preserve individual research integrity and protect against improper influence.

Even for those working in independent laboratories, attention must be directed toward cases submitted by law enforcement or other agencies to limit bias. Some anthropologists insist that no contextual information be provided about the case being analyzed to limit any possibility of bias. More commonly, anthropologists will be provided only the information they need to know to properly analyze the forensic aspects of the case. The analyst must discourage any investigator who tries to convey theories or desired research results.

Most of my casework experience involved remains submitted by the FBI laboratory. From the beginning of my work with the agency we had an understanding that I would be provided only the basic information that I needed to know with each case. I have submitted numerous independent reports and cannot recall any instance of attempts to improperly influence my ana­lysis. When I have worked with other individuals and organizations on forensic issues, I have always made it a point to indicate at the outset my requirements to preserve objectivity.

I recall one case in particular involving human remains the FBI submitted to me for analysis. Although I found evidence of damage to the remains, it was not clear whether the alterations represented taphonomic factors sustained after death or if they could be linked to the death event itself. Years after submitting my report, I learned that local prosecutors were bringing charges against an individual, arguing that the altera­tions I had observed were produced by sharp force trauma leading to the death of the victim. Since my report did not conclude that sharp force trauma was involved, I was summoned to the pre-trial hearing by the defense. This case reveals that although the expert analysis may be requested by a law enforcement agency, the results may ultimately benefit the defense effort.

When considering employment with a new organization, the forensic anthropologist should discuss these kinds of issues. Consider the goals of the organization, especially in the global arena of human rights investigations. Can the integrity and objectivity of the forensic anthropologist be protected within the structure and goals of the organization? In most cases, these issues can be addressed but only if they are recognized early and explicitly discussed. Failure to secure proper safeguards can later produce tension and even conflict. In some circumstances, it may be useful to produce a document clarifying concerns and the safeguards that need to be constructed. Payment for forensic consultation is proper but should never be arranged on a contingency basis.

While most factors of potential bias can be recognized and limited, others may be more subtle. To limit cognitive bias, the anthropologist must safeguard against easily made assumptions that may be incorrect. These can relate to status as a criminal case, time since death, association with artifacts, elements of the biological profile and most other areas of anthropological analysis. At trial, the expert must not assume guilt or innocence of the defendant. While questions posed at trial may reveal the perspective of the lawyer, the expert’s answers must be anchored in objectivity. Even if the lawyer at trial is challenging the witness’ credi­bility and may try to lead the expert to give a particular response, it is important to try to stay objective and neutral. Some technology can reduce the chances of bias but does not completely prevent it. Sources of bias in addition to those discussed above include training and personal factors relating to the analyst and basic human nature [ 8 ].

Objectivity must also be protected in the research process. Research involves formation of a hypothesis and testing. The research design should ensure objective data collection and analysis that fairly tests the hypothesis. Analysis needs to include proper statistical treatment and reporting of results with appropriate indications of the probabilities and errors involved. In both research and case reports, the conclusions must be consistent with the data and/or bench notes.

Efforts should be made to limit typographical mistakes and other errors in case reports and research manuscripts. Wording should be carefully considered to accurately reflect content and intent. Years ago, I developed a practice of repeatedly proof-reading docu­ments with a co-author or colleague until no errors could be detected and no further word adjustments were required. Usually, this requires multiple proof-reading sessions. References should be drawn from original sources and arranged in the correct format of the intended publication or other targeted outlet.

Collections

Forensic anthropology research is largely collection based. To develop and test methods and hypotheses, forensic anthropologists frequently rely on documented collections of human remains and related materials. Such documentation includes individual information relating to age, sex, living stature, ancestry, disease history, and other such variables. Collections usually consist of skeletonized human remains but can include images and similar clinical information. Research use of such collections involves ethical considerations and judgment on a variety of issues [ 9 ].

The acquisition of new collections and perhaps even the use of collections raises questions that require examination of legal and ethical issues. Were the remains legally acquired? What is the documentation based upon and what is the probability of accuracy? Was there proper informed consent in the acquisition? Are the remains subject to repatriation laws and/or institutional policies? What permissions are required to access remains and conduct research? Once research is conducted, what is the proper manner to report/disseminate results?

Research integrity related to collections use involves proper safety measures, minimizing damage to the collections and proper collection selection. Adequate protection must be provided in the field, laboratory and storage units against fire, flood, high wind, theft, and other similar threats. Safety is usually a minimal concern when working with skeletonized museum collections. Proper masking for dust inhalation protection and wearing protective gloves are common protective measures. More advanced protection may be required when working with fleshed remains or those in an advanced state of decomposition [ 10 ].

Even routine handling of collections can be destructive when bones and teeth are in contact with each other and with hard surfaces. Efforts should be made to minimize direct contact and, when possible, guard against contact with hard surfaces, using soft surface coverings, cork rings for placement of crania, etc. Care must also be taken in placing remains in storage containers, ensuring that remains are not damaged when applying lids or placing units into storage structures. The storage environment should be regulated for proper temperature and humidity control. In some cultural contexts, community consultation as far as treatment of remains may be appropriate.

Although obvious to most researchers, mention is due to proper selection of collections. The research process calls for hypothesis testing and choosing collections that are appropriate for the research design [ 11 ]. For example, testing a method of estimating age at death calls for use of a documented collection of individuals of known age, not a collection recovered from an archaeological context with only estimated ages. Proper collection selection can involve regional considerations, sample size, preservation issues and availability.

Considerable recent research [ 12 ] involves destructive analysis. Histological research, the developments of isoscapes, DNA analysis and many other new approaches all call for various forms of destructive sampling. For such sampling, proper permissions must be obtained, and destruction should be kept as minimal as possible. Institutional guidelines must be followed regarding documentation prior to sampling and disposition of residual material. If multiple samples are involved, care must be taken to limit inter-sample contamination. Safeguards against environmental contami­nation are especially important in molecular research.

Research and casework in forensic anthropology call for using the best and most appropriate methods avai­lable. Such use includes appropriate statistical analysis, proper incorporation of technology, and thoughtful research design. Conclusions must be consistent with results and accurately reflect the limitations of the methods used. Conclusions and reports must present correct levels of probability and error assessment to ensure trust in the results.

Some consideration must also be given to the involvement of families and communities in research endeavours. In recent years, the global anthropological community has moved toward greater inclusion of such perspectives in research, casework and issues relating to collections. Such perspectives can be essential in the acquisition, curation, research use and repatriation of collections. Forensic applications related to human rights and humanitarian action routinely involve the affected community and family-based organizations [ 13 ]. Such involvement reflects shared interests in the research process and can provide valuable insights. Of course, forensic science must also guard against improper influence and bias and seek to preserve objectivity.

Training in forensic anthropology can be challenging. Forensic anthropologists address a broad range of legal issues involving a multitude of methods and problems to solve. The field of forensic anthropology includes participation in search and recovery, excavation, docu­mentation, species determination, assessment of sex, age at death, time since death, living stature and ancestry, evaluation of taphonomic alterations, assessment of trauma and evidence of foul play and identification. The forensic anthropologist must be familiar with legal issues and courtroom procedures [ 14 ] and applications in related areas of forensic science. Analysis includes not only the traditional calipers, but also advanced statistics and use of computerized databases. Specialized techniques may be called for, such as histology, isotope analysis, radiocarbon analysis, facial approximation and craniofacial photographic superimposition. Those working internationally need to understand law and policy relating to humanitarian action and the investigation of human rights abuse [ 15 ]. Each of these areas of forensic anthropology practice presents a substantial field of published literature, history of application and complex methodology.

The well-trained forensic anthropologist needs to be competent in all the basic areas of application and at least have a working knowledge of the value of the specialized techniques [ 16 ]. University graduate programsme face challenges in providing adequate training in such diverse academic areas. Students entering these programmes must decide on a programme of course work and research that will provide the necessary background. Many of the training programs are anchored within university anthropology departments that also require coursework in social anthropology, linguistics, and archaeology. In addition, the forensic anthropology graduate student may desire coursework in other departments focusing on anatomy, statistics, and other areas of forensic science. Advanced training in writing, public speaking, grant preparation, finance management and report structuring can also bolster confidence and enhance career development. This vast array of options forces hard decisions on the academic path to follow. Integrity involves choosing the appropriate path given the opportunities available and career goals and then honestly projecting that training in forensic applications. For example, the forensic anthropologist with no training in excavation should reveal that lack of experience when presented with a recovery operation requiring such expertise. Mentors can also play a criti­cal role in guiding students as they navigate career decisions and understand the value of proceeding with integrity regardless of their chosen path.

In casework, or in any domain, the forensic anthropologist needs to recognize his/her academic limitations and disclose them, guarding against offering opinions beyond their expertise or in other areas of forensic science. A forensic anthropologist may be well qualified to render opinions on dental anatomy and dental age changes but should leave analysis of dental restorations to the forensic odontologist. Forensic anthropologists may find the evidence of trauma, but in most jurisdictions, forensic pathologists must determine cause and manner of death. This need for humi­lity and transparency also applies to clothing analysis, ballistics, soft tissue interpretation, legal issues and other areas of forensic science, or any field for that matter, in which other specialists are better positioned academically to render opinions.

Accurate recognition of training and expertise provides the foundation for organizational ethics requirements. For example, the code of ethics and conduct of the American Academy of Forensic Sciences (AAFS) states that an AAFS member or affiliate must not “materially misrepresent his or her education, training, experience, area of expertise, or membership status within the Academy”, as well as “data or scientific principles upon which his or her conclusion or professional opinion is based” [ 17 ]. While it may be easy to embrace such laudable declarations, some subtle nuances can be challenging, especially when under the duress of court testimony. Personal integrity can dictate the path forward in these situations.

Authorship, plagiarism and peer review

Publications and final case reports represent central goals in forensic anthropology practice. Given their importance in the research process, they deserve special attention in efforts to preserve integrity. Authorship issues top the list of areas of concern. The author list should include all individuals who made intellectual contributions and exclude those who did not. The journal Forensic Sciences Research facilitates this decision-making process by requesting that upon manuscript submission, the contributions of each author be defined (available from: https://www.tandfonline.com/action/authorSubmission?show=instructions&journal-Code=tfsr20, accessed 25 July 2021). The Journal of Forensic Sciences also provides guidance on authorship, noting that data collection and general supervision of the research group do not qualify (available from: https://onlinelibrary.wiley.com/page/journal/15564029/homepage/forauthors, accessed 25 July 2021). The first author should be the person with primary responsibility for the manuscript content. Most journals agree that the order of authors listed should be decided and agreed upon by the authors themselves. The International Committee of Medical Journal Editors (ICMJE) offers useful guidelines on authorship and related scholarly publishing issues (available from: http://www.icmje.org/recommendations, accessed 25 July 2021).

Plagiarism represents a major integrity concern for authors and journals alike. As defined in the US regu­latory context, plagiarism consists of the “appropriation of another person’s ideas, processes, results, or words without giving appropriate credit” (available from: https://ori.hhs.gov/definition-research-misconduct, accessed 25 July 2021). This definition has been incorporated or adapted in codes and guidelines around the world. In essence, scholarly work must be original and not published elsewhere without disclosure and permission. If detected, plagiarism can land a near fatal blow to an individual’s integrity. Many journals use plagiarism detection software, and some declare the right to bar an author from future publication if plagiarism is detected. Authors must take care to cite the work of others and seek permissions to publish direct quotes of major length, as well as illustrations, whether published or presented. Scientific writing must exemplify respect and acknowledgment, giving proper credit to others whenever appropriate to do so. For authors who publish extensively on similar topics, self-plagiarism can emerge as an issue. Subtle self-plagiarism can usually be avoided with original text, illustrations, and tables, but concerned scholars can consult detection systems such as www.ithenticate.com .

Most journals also require disclosure of funding sources and ethics compliance in the use of human or animal subjects. Citing funding sources (and other interests that could give rise to a real or perceived bias) provides credit to those supporting projects but also reveals any potential influence of the source on the project goals and results. The ethics statements serve as useful reminders that rules may be in place for some projects and likely vary with different institutions.

Publishing trustworthy results in reputable journals enhances research integrity. In recent years, peer review has increased as an issue for scholars’ career advancement in forensic science. In most cases, the scholar’s home institution considers the extent of peer review in evaluating the scholar’s publication record. The impact factor of the journal has also grown in importance in this regard. Many journals in forensic science have formed recently, presenting a variety of publishing options but also highly variable systems of peer review and business approach. Selecting the most appropriate journal for publishing research results can be guided by an examination of each journal’s editorial board in addition to the journal’s impact factor. Advice from respected senior colleagues can also prove especially useful.

Some colleagues cringe at the idea of peer review, largely due to the often extensive amount of time involved and the fear of unwarranted critical review. Although at times I have shared the concern about the length of time required for review, I generally welcome peer review. Reputable journals send review articles to esteemed colleagues who usually offer criti­cal but positive advice on how to improve a manuscript. My own work has almost always been improved through the thoughtful suggestions of colleagues in peer review. Outside perspectives from a knowledgeable colleague should be welcomed and appreciated.

Integrity issues abound in relation to scientific publishing. Once receiving critical review, authors usually either respond positively with editing or defend their point of view. In cases of manuscript rejection or severe criticism, some authors may seek publication in another venue. In other scenarios, an author may have participated voluntarily in an edited volume or meeting symposium and then, following criticism, chooses to publish the contribution singly elsewhere. In such cases, is the author obligated to acknowledge the previous criticism? Stewart [ 18 ] raised this issue as early as 1946, revealing that few guidelines were available at that time. Today, many journals request permission of reviewers to transmit their comments to another journal if the manuscript is submitted elsewhere following review and rejection.

In sum, issues inherent to responsible authorship and publication, avoiding plagiarism, and engaging in peer review present vibrant opportunities for demonstrating one’s integrity in the context of forensic research.

A commitment to integrity in training also involves giving back. Ideally, experienced forensic anthropologists share their expertise with students and prospective forensic specialists through formal training programmes, lectures, workshops, and advice. Involvement of students in casework and research represents a prime teaching opportunity. Such activity can be symbiotically beneficial to both the instructor and the students, provided that the experience is appropriately structured and clear expectations and safeguards are in place. Care must be taken to ensure that students do not engage in forensic activities beyond their expertise or violate any policy relating to casework.

Sharing expertise, both formally and informally, is particularly important in global casework and research. A forensic specialist may be invited to analyze and report on a particular case in a country different from their own. That experience can be rewarding and may contribute to the resolution of the case. The impact of his or her involvement, however, can be increased if the casework is supplemented by local capacity building through workshops or direct training in case analysis. Most specialists involved in global capacity building report learning as much as they teach through such engagement. Impact is greatly enhanced if relationships are forged and the specialist’s visit leaves local colleagues trained to bring improved forensic practice to bear on future cases. Given the global interaction made possible by the modern internet, international sharing of expertise, knowledge, methods, and data have become not only relatively common but essential in some cases and nearly always rewarding.

Summary and conclusions

While integrity in forensic research remains a universal goal, it is difficult to precisely define. The above discussion reveals key elements of research integrity and offers suggested guidelines. However, the individual forensic anthropologist must find their own way, given their training and work environment, as they navigate the ethical issues involved.

Modern forensic anthropologists need to preserve objectivity in their research and casework and take measures to limit bias. In report writing and court testimony, the forensic anthropologist must stay within their domain of expertise and avoid misrepresentation of their credentials and data. Effort should be made to limit errors in report writing and publications. Ethics statements of forensic organizations provide useful guidelines and should be shared and discussed with students.

Integrity in forensics research includes proper use of collections, especially regarding safety issues. Effort must be made to limit and carefully select any destructive analysis. Collection conservation procedures must be followed to limit damage to collections through use. Appropriate selection of collection samples for research must be made to match the research goals. Similarly, thoughtful selection of research methods contributes to research integrity.

Authorship deserves special attention, giving proper credit to those intellectually engaged in the project and writing. Plagiarism, including self-plagiarism, must be avoided through careful citation, and obtaining needed permissions. Forensic scholars must value peer review, giving proper responses if being reviewed and being fair if reviewing the work of others.

Community involvement and culturally sensitive communication represent important components of both research and casework, especially in global human rights work. Whenever possible, training and mentoring others and capacity building should be encouraged as important goals.

The forensic anthropologist conducting research and casework with integrity enjoys substantial professional rewards. Conducting a successful research process culminating in publication in a peer-reviewed journal or book represents a joyful professional experience. Such research establishes new methods, tests existing methods, and brings new technology to bear on looming forensic problems. Activity in forensic anthropology offers the reward of helping to resolve key issues of contemporary society.

Fortunately, many gifted and highly qualified students are attracted to the field of forensic anthropology. Today, these prospective forensic anthropologists find many educational opportunities and a growing, favourable employment environment. As they explore these opportunities and gain expertise, they must also focus on building credibility and maintaining integrity, keeping abreast of emerging technologies and ethical issues while remaining true to their personal and professional values.

Forensic Anthropology Center

• Affiliated Faculty/Researchers
• Human Remains Recovery Course
• Forensic Anthropology Methods
• Forensic Entomology
• Human Osteology
• K9 Human Remains Detection
• Identifying Human vs. Nonhuman Bone
• TEEX Skeletal Death Investigation Course
• Forensic Art: Forensic Facial Reconstruction Sculpture
• Forensic Art: Drawing to Depict the Deceased for Identification
• Competency/Proficiently Assessment
• Applied Biostatistics Using R
• Forensic Anthropology and Osteological Analyses
• Operation Identification
• Researchers
• Forensic Anthropology Research Facility
• Osteology Research and Processing Lab
• Grady Early Forensic Anthropology Laboratory
• Laboratory SOP
• Graduate Courses
• Frequently Asked Questions
• FACTS Summer Scholar Program
• Whole Body Donation
• FARF Research Request Form
• CT Service Request Form
• Bone Sampling Request Form
• Skeletal Collection Research Request Form
• Current Students
• Faculty & Staff
• Family & Visitors

Information on whole body and financial donations

Upcoming workshops | TCOLE and ABMI credits available

Research and Professional Services

Information for Scientific Researches

Community Outreach, Talks and Tours, Case Work