Papers by Kaleigh C Best
Journal of forensic sciences, 2018
When faced with commingled remains, it might be assumed that a more "masculine" pelvis ... more When faced with commingled remains, it might be assumed that a more "masculine" pelvis is associated with a more "masculine" cranium, but this relationship has not been specifically tested. This study uses geometric morphometric analyses of pelvic and cranial landmarks to assess whether there is an intra-individual relationship between the degrees of sexual expression in these two skeletal regions. Principal component and discriminant function scores were used to assess sexual dimorphism in 113 U.S. Black individuals. Correlation values and partial least squares regression (PLS) were used to evaluate intra-individual relationships. Results indicate that the os coxae is more sexually dimorphic than the cranium, with element shape being more sexually dimorphic than size. PLS and correlation results suggest no significant intra-individual relationship between pelvic and cranial sexual size or shape expression. Thus, in commingled situations, associations between the...
Journal of Forensic Sciences, 2017
When faced with commingled remains, it might be assumed that a more "masculine" pelvis is associa... more When faced with commingled remains, it might be assumed that a more "masculine" pelvis is associated with a more "mascu-line" cranium, but this relationship has not been specifically tested. This study uses geometric morphometric analyses of pelvic and cranial landmarks to assess whether there is an intra-individual relationship between the degrees of sexual expression in these two skeletal regions. Principal component and discriminant function scores were used to assess sexual dimorphism in 113 U.S. Black individuals. Correlation values and partial least squares regression (PLS) were used to evaluate intra-individual relationships. Results indicate that the os coxae is more sexually dimorphic than the cranium, with element shape being more sexually dimorphic than size. PLS and correlation results suggest no significant intra-individual relationship between pelvic and cranial sexual size or shape expression. Thus, in commingled situations, associations between these skeletal elements cannot be inferred based on degree of "masculinity."
Conference Presentations by Kaleigh C Best
This presentation will impact the attendees by providing classification accuracies for a traditio... more This presentation will impact the attendees by providing classification accuracies for a traditionally ill-represented, unique ancestral population using modern statistical techniques. Skeletal biologists will be able to explore the variation present in the Chinese-Cuban sample and how it compares to known modern and historic reference samples.
Fordisc is a widely used tool for the estimation of ancestry based off on a number of cranial measurements that differ between population groups. Cuba has a diverse genetic background that includes Native American, African, Spanish, and Chinese influences. Native American groups, decimated and enslaved by Spanish, eventually dwindled to numbers that lead to the importation of slaves from West Africa. After the banning of slavery in the late 1800s, Chinese were brought to the island as indentured servants and later middle-class Chinese businesspersons came to the island in the 1920s. These waves of migrants from three continents helped create the diverse Cuban population that is present today. Remains from a recent Chinese cemetery exhumed and now incorporated into the Aristides Mestre Laboratory of the Montanė Anthropology Museum, University of Havana, Cuba, includes 49 Chinese males, according to records. It is the goal of this project to investigate how this group classifies using Fordisc.
Fourteen cranial measurements, as defined in Standards, were collected by CW from the 49 Chinese-Cuban males. Individuals were then analyzed in Fordisc 3.11 using the Forensic Data Bank (FDB) and Howells groups in accordance with the procedures outlined in Ousley and Jantz 2. The posterior probabilities and typicality probabilities were then utilized in eliminating groups that did not meet acceptable cutoffs (posterior probabilities > 0.1 and F-typicality > 0.05). Classification accuracies were evaluated using leave-one-out cross-validation. Further analyses included comparing the Cuban male sample as a whole to forensic and Howells groups using the Mahalanobis distance.
Using the FDB groups, individuals classified as Chinese 32.7% and as an Asian group 63.3% of the time. A further breakdown of classifications of groups is: Vietnamese Male=16.3%, Japanese Male = 14.2%, Guatemalan Male = 14.2%, Black Male= 8.2%, American Indian Male= 6.1% and Hispanic Male =2.0%. One individual classified as a Japanese female and two individuals could not be classified due to low typicality scores. The correct sex classification rate for the sample was 98%. When analyzed as a group using Fordisc, the linear discriminant function classified 32.7% (16/49) of the sample into its own group when BPL was removed due to the large number of the sample missing this measurement. Results demonstrate that the Wienker sample is closest to Asian groups, particularly the Chinese males, with a Mahalanobis distance of 2.5. Other close groups include the Vietnamese males, with a Mahalanobis distance of 3.3, and the Japanese males, with a Mahalanobis distance of 3.6. The Cuban-Chinese males were most similar to Howells' Chinese groups (Anyang and Hainan), followed by other East Asian groups.
In summary, classification results corroborated the recorded Chinese origins of the sample. Individuals from this sample classified as East Asian 63% of the time. Even when attributed to a single ancestry, the Chinese-Cubans classify as Chinese males 33% of the time, the highest classification rate of the groups compared. These classification percentages are better than random chance and thus indicate that a large component of the morphological variation reflects the group's geographic source. As a group, classifications of the sample using FDB and Howells data reflect Southeast Asian affinities, particularly indicative of Chinese heritage. The comparative samples come from different areas in China and reflect different ethnic groups. The Fordisc Chinese sample includes individuals from Hong Kong, and the Howells groups include the Anyang from northern China and the Hainan from southern China.
This study utilizes three female adult pigs approximately between 150-200 lbs. as human proxies.
The os coxae and the cranium are often mentioned as the most sexually dimorphic regions of the hu... more The os coxae and the cranium are often mentioned as the most sexually dimorphic regions of the human skeleton 1 , and thus are often used to estimate the sex of individuals in a variety of physical anthropology subfields, including paleoanthropology, bioarchaeology, and forensic anthropology. Sexual differences in these regions have been attributed to differential hormonal expression 2-4 , functional or biomechanical constraints 5-9 , and sexual selection/mating preferences 9. However, the direct influence of each of these variables on skeletal sexual dimorphism remains relatively unexplored. Further intra-individual levels of the degree of sexual dimorphism expressed have not been studied. Indirectly, Stull and collegues 10 compared non-metric sex classification accuracies of the Klales et al. 11 and Walker 12 traits of 112 individuals. Although not directly comparing intra-individual levels of dimorphism, their results show that the ossa coxae were more accurate in sex estimation than the cranium and therefore suggests that the sexual expression in these two skeletal regions is influenced by different factors (or if they are influenced by the same factors, there is a difference in the degree of influence). Only one other study looking at sexual dimorphism, by Schutz et al., 13 examines size and shape of the cranium and pelvis within two species of foxes. Results indicated that no significant sex differences in the fox crania existed, but the ossa coxae were dimorphic in both species. The authors attribute these differences to the reproductive life-histories and mode of locomotion required of canids, both of which would have a higher influence on the pelvis than the cranium. Research Objective The aim of this study was to compare intra-individual levels of cranial and pelvic sexual expression. If the same factors (e.g., hormones) influence sexual dimorphism in both the cranium and pelvis, then significant correlations should be expected between the levels of sexual expressions of these two skeletal regions. A sample of 113 ossa coxae and crania collected from U.S. Blacks in the Hamann-Todd Osteological Collection were analyzed (Table 1). Table 1: Sample distribution Forty-two landmarks from the cranium and 12 landmarks from the left os coxae (Figure 1) were digitized using a MicroScribe ® digitizer. Landmarks were chosen in order to capture the overall shape and size of the skeletal elements. For the crania, landmarks were collected following definitions provided in the FORDISC 3.1 help file (see Jantz and Ousley 14 for more information). Landmarks for the ossa coxae were collected using guidelines from various authors (see Table 2) 15, 16. Following a Procrustes superimposition, principal component (PC) and cross-validated discriminant function analyses (DFA) were used to assess and compare the degree of sexual shape and form dimorphism present in both skeletal regions. Centroid size was used to assess sexual size dimorphism. Spearman's rank and Pearson's correlations were then calculated (using both pooled and sex-specific samples) between the pelvic and cranial discriminant function scores obtained for each individual in order to assess whether there is a significant intra-individual relationship between the degree of sexual expression in these elements. Partial least squares regression was also performed on the pelvic and cranial PCs and Procrustes coordinates in order to assess the relationship between these two blocks of data. Sex-specific Spearman's rank and Pearson's correlation tests revealed no significant relationship between the os coxae and the cranium discriminant function scores (Table 3, Figure 2). In other words, males and females with a greater pelvic masculinity do not necessarily have similarly ranked levels of cranial masculinity. Partial least squares (PLS) revealed no significant relationship between the os coxae and the cranium whether run on the Procrustes coordinates (RV coefficient = 0.09) (Figure 3) or PCs (RV coefficient = 0.08) (Figure 4). Centroid Size and Form Analyses on form data (where size was retained in the analyses) did not change the results; no significant relationships were observed between pelvic and cranial variables. Similarly, there was no significant relationship between centroid size values of the os coxae and cranium.
After attending this presentation, attendees will understand the effects of prior freezing on hum... more After attending this presentation, attendees will understand the effects of prior freezing on human decomposition and whether frozen human corpses display the suite of decompositional traits previously identified in frozen pig proxies.1 This presentation will impact the forensic science community by providing preliminary answers to the questions of how previous freezing impacts the rate and pattern of human decomposition and whether there are visible external features that may identify previously frozen human corpses during decomposition, as were identified in pig proxies. Freezing is an important forensic taphonomic variable as subjects may be stored in the frozen condition prior to use in taphonomic studies. Additionally, perpetrators may freeze victims in attempts to delude investigators by complicating Postmortem Interval (PMI) estimation. Freezing may impact decomposition by decreasing the viability of enteric bacteria responsible for driving putrefaction and research has shown that previous freezing significantly decreases the rate of decomposition in pigs.1-3 In line with recent calls to validate results of forensic taphonomic studies using animal models with follow-up research utilizing human subjects, this study investigated human frozen decomposition.4,5 It was hypothesized that the decrease in viable bacteria after freezing would result in a slower rate of decomposition in frozen human subjects, and human and pig subjects would show similar externally visible effects of previous freezing. All subjects in this study were deposited at the Complex for Forensic Anthropology Research (CFAR) at Southern Illinois University. To understand how previous freezing alters the progression of “normal” human decomposition, two comparisons were made. A previously frozen human subject deposited on September 18, 2015, the fall season in southern Illinois, was compared to a never-frozen human control deposited on the same date and to never-frozen human subjects deposited in the fall of 2012-2015 (n=5). Additionally, the pattern of human frozen decomposition was compared to previously reported results from a similar study using pig proxies to identify whether the markers of decomposition observed in frozen pigs are identifiable in frozen humans as well. One adult human subject was frozen at -18°C for 172 days. Fresh subjects were deposited immediately upon arrival at CFAR. Human donors arrived at CFAR within eight days of death and were kept in morgue refrigeration prior to deposition. All subjects were placed nude, directly on ground surface, within chain-link enclosures to minimize scavenging. Total Body Score (TBS), abdominal circumference, digital images, and written observations concerning insect activity, scavenging, and weather were collected daily.6 To quantify decomposition rate, Kelvin Accumulated Degree Days (KADD) were used to measure the thermal energy required for each subject to reach several TBS thresholds: early decomposition (TBS≥6.0); halfway through early decomposition (TBS≥12.5); advanced decomposition (TBS≥19.0); halfway through advanced decomposition (TBS≥23.0); and skeletonization (TBS≥27.0). Bloat was quantified by analyzing the percent difference in abdominal circumference for each subject. The previously frozen human subject decomposed faster than its paired never-frozen human control during both early and advanced decomposition. Percent difference in abdominal circumference was 105% for the frozen subject compared to 123% for the control. Bloat in the control subject was more rapid, substantial, and sustained. In addition to minimal bloat, the previously frozen subject displayed several differences in decomposition pattern, including grayer, pasty overall color, an absence of green discoloration, earlier onset of bone exposure, and a faster deflation of tissues compared to the control subject. With the exception of earlier onset of bone exposure, differences in decomposition pattern were consistent with previous frozen pig decomposition research. Compared to never-frozen human subjects deposited in the fall at CFAR, the previously frozen subject fell within the range of thermal energy necessary to achieve early and advanced decomposition, but at TBS≥23.0 and TBS≥27.0, the frozen subject decomposed more rapidly. In this preliminary investigation, prior freezing was found to increase the rate and alter the pattern of human
81 *Presenting Author
decomposition. Additionally, the rate of frozen human decomposition did not progress as expected based on previous research on frozen pig decomposition, but expected effects on visible external features were observed. Further research is ongoing.
After attending this presentation, attendees will understand how sexual size and shape dimorphism... more After attending this presentation, attendees will understand how sexual size and shape dimorphism compares in the human os coxae and cranium, including which shape changes contribute the most to sex differences. This presentation will impact the forensic science community by providing cross-validated correct sex classification rates for the os coxae and cranium and demonstrating to forensic anthropologists which of the two elements provides the most reliable sex estimation when case results may be contradictory. In humans, the os coxae and the cranium are commonly referred to as the most sexually dimorphic regions of the skeleton and thus are often used to estimate the sex of individuals in a variety of physical anthropology subfields, including paleoanthropology, bioarchaeology, and forensic anthropology. Although there are numerous studies analyzing either pelvic or cranial sexual dimorphism, these studies utilize various populations, samples, sample sizes, types of data, methods, or statistical analyses, making a direct comparison between the resultant dimorphism values invalid. Only one study was found in which pelvic and cranial dimorphism was analyzed using a single sample, but it only compared non-metric sex estimation results, which are known to be somewhat subjective. The goal of this study was to use landmark data and geometric morphometric analyses to compare sexual size and shape dimorphism in the os coxae and cranium in a single sample. The use of a single study sample for both analyses means that the obtained sex classification results are directly comparable and will provide forensic anthropologists with information regarding the reliability of these two elements in sex estimation methods. Forty-two landmarks from the cranium and 12 landmarks from the os coxae were digitized using a MicroScribe® on a sample of 113 United States Black adults (aged 17 years to 70 years) from the Hamann-Todd Osteological Collection. Following a Procrustes superimposition, principle component and discriminant function analyses were used to assess and compare the degree of sexual shape and form (combined shape and size) dimorphism present in both skeletal regions. Univariate analyses were performed to evaluate which specific shape changes were contributing the most to the sex differences. Centroid size was used to assess sexual size dimorphism. The results of the shape, form, and size analyses all indicate that the os coxae is more sexually dimorphic than the cranium. The discriminant function analysis performed on the os coxae shape variables resulted in a cross-validated correct sex classification rate of 99.1%, compared to 84.1% in the cranium. Including size in the shape analyses (i.e., form) and analyzing size independently (i.e., centroid size) did not increase sex classification rates, indicating that sex differences in these elements occurs primarily in shape. The geometric morphometric analyses confirmed that relative pubis length, sciatic notch breadth, and subpubic concavity were the most important shape variables in pelvic dimorphism. In addition, os coxae height relative to ilium breadth also contributed to sex differences. In the cranium, the geometric morphometric analyses revealed sex differences in facial height, vault breadth, cranial base flexion, nasal width, and glabellar prominence. As this study uses a single sample to analyze sexual dimorphism in the os coxae and cranium, it eliminates many extraneous variables (e.g., sample and method differences) and allows a direct comparison between the skeletal regions. Results confirm that the os coxae is significantly more dimorphic than the cranium; thus, when assigning sex to an unknown skeleton, forensic anthropologists should rely more heavily on pelvic morphology. Geometric morphometric shape analyses conducted on the os coxae landmark data can discriminate between the sexes with up to 99% accuracy and provide an objective method to quantitatively analyze traditional non-metric sex traits
Book Reviews by Kaleigh C Best
Midcontinental Journal of Archaeology, 2019
encourages the reader to critically evaluate the ethics that are involved in the use and curation... more encourages the reader to critically evaluate the ethics that are involved in the use and curation of identified skeletal collections, defined as those collections of human remains derived from "archaeological sources, dissections and other cadaver sources, and those from recent cemetery sites" (p. 1.). While biological and forensic anthropologists routinely rely on identified skeletal collections for both research and applied means, this volume challenges the anonymous treatment of individuals within these collections and emphasizes the underlying biases inherent in all collections. Beyond highlighting these foundational issues, the authors then elucidate the many facets of these biases and suggest ways that professionals can navigate through them with benefit to the discipline. Finally, this volume fosters discussion over the ethics of using identified skeletal collections and how researchers should engage with the individuals interred within the collections as well as with descendent communities.
Uploads
Papers by Kaleigh C Best
Conference Presentations by Kaleigh C Best
Fordisc is a widely used tool for the estimation of ancestry based off on a number of cranial measurements that differ between population groups. Cuba has a diverse genetic background that includes Native American, African, Spanish, and Chinese influences. Native American groups, decimated and enslaved by Spanish, eventually dwindled to numbers that lead to the importation of slaves from West Africa. After the banning of slavery in the late 1800s, Chinese were brought to the island as indentured servants and later middle-class Chinese businesspersons came to the island in the 1920s. These waves of migrants from three continents helped create the diverse Cuban population that is present today. Remains from a recent Chinese cemetery exhumed and now incorporated into the Aristides Mestre Laboratory of the Montanė Anthropology Museum, University of Havana, Cuba, includes 49 Chinese males, according to records. It is the goal of this project to investigate how this group classifies using Fordisc.
Fourteen cranial measurements, as defined in Standards, were collected by CW from the 49 Chinese-Cuban males. Individuals were then analyzed in Fordisc 3.11 using the Forensic Data Bank (FDB) and Howells groups in accordance with the procedures outlined in Ousley and Jantz 2. The posterior probabilities and typicality probabilities were then utilized in eliminating groups that did not meet acceptable cutoffs (posterior probabilities > 0.1 and F-typicality > 0.05). Classification accuracies were evaluated using leave-one-out cross-validation. Further analyses included comparing the Cuban male sample as a whole to forensic and Howells groups using the Mahalanobis distance.
Using the FDB groups, individuals classified as Chinese 32.7% and as an Asian group 63.3% of the time. A further breakdown of classifications of groups is: Vietnamese Male=16.3%, Japanese Male = 14.2%, Guatemalan Male = 14.2%, Black Male= 8.2%, American Indian Male= 6.1% and Hispanic Male =2.0%. One individual classified as a Japanese female and two individuals could not be classified due to low typicality scores. The correct sex classification rate for the sample was 98%. When analyzed as a group using Fordisc, the linear discriminant function classified 32.7% (16/49) of the sample into its own group when BPL was removed due to the large number of the sample missing this measurement. Results demonstrate that the Wienker sample is closest to Asian groups, particularly the Chinese males, with a Mahalanobis distance of 2.5. Other close groups include the Vietnamese males, with a Mahalanobis distance of 3.3, and the Japanese males, with a Mahalanobis distance of 3.6. The Cuban-Chinese males were most similar to Howells' Chinese groups (Anyang and Hainan), followed by other East Asian groups.
In summary, classification results corroborated the recorded Chinese origins of the sample. Individuals from this sample classified as East Asian 63% of the time. Even when attributed to a single ancestry, the Chinese-Cubans classify as Chinese males 33% of the time, the highest classification rate of the groups compared. These classification percentages are better than random chance and thus indicate that a large component of the morphological variation reflects the group's geographic source. As a group, classifications of the sample using FDB and Howells data reflect Southeast Asian affinities, particularly indicative of Chinese heritage. The comparative samples come from different areas in China and reflect different ethnic groups. The Fordisc Chinese sample includes individuals from Hong Kong, and the Howells groups include the Anyang from northern China and the Hainan from southern China.
81 *Presenting Author
decomposition. Additionally, the rate of frozen human decomposition did not progress as expected based on previous research on frozen pig decomposition, but expected effects on visible external features were observed. Further research is ongoing.
Book Reviews by Kaleigh C Best