The Efficacy of Traditional and Digitally-Derived Anthropometry Among Black Women

Material Information

The Efficacy of Traditional and Digitally-Derived Anthropometry Among Black Women
Osbourne, Brittany
Place of Publication:
Gainesville, Fla
College of Fine Arts; University of Florida
Publication Date:
Physical Description:
Project in lieu of thesis


Subjects / Keywords:
Anthropometry ( jstor )
Apples ( jstor )
Avatars ( jstor )
Biomedical technology ( jstor )
Body mass index ( jstor )
Fats ( jstor )
History of technology ( jstor )
Obesity ( jstor )
Waist ( jstor )
Women ( jstor )


This project tests the efficacy of handheld technology against traditional direct methods to inexpensively gather anthropometric data from African American and Afro-Caribbean women living in Florida, and to help create an ethnic-specific digital database for future anthropometric survey use. Results demonstrate that indirect anthropometry that uses handheld technology has a shorter examination period, but provides less precision. This proposed project also explores the benefits and drawbacks of replacing 2D body image charts in clinical research and practice with charts derived from anthropometry that uses handheld 3D technology. How participants interact with and experience handheld technology, while undergoing anthropometric examinations, are also reported. Based on methodological validity, direct anthropometry is more effective. More participants also prefer the direct anthropometric exam as opposed to the indirect methods derived from digital technology. However, although in its prototypical phase, the technology used during this project presents unique opportunities to compare individual body shape and size, which may serve as a powerful medical screening tool in the future.
General Note:
Digital Arts and Sciences terminal project

Record Information

Source Institution:
University of Florida Institutional Repository
Holding Location:
University of Florida
Rights Management:
All rights reserved by the source institution and holding location.
Resource Identifier:
915054047 ( OCLC )


This item is only available as the following downloads:

Full Text




2 2013 Brittany Osbourne


3 To my intellectual grandmother who worked from the intersection of anthropology and the fine arts, and was unafraid to stand on a street corner in Harlem, take a pair of calipers, and measure anyone who passed by. With her pencil and caliper, she fought against pseudoscientific racism. Zora Neale Hurston, I dedicate this project to you.


4 ACKNOWLEDGEMENTS I am extremely thankful for the women who allowed me into their homes to conduct anthropometric examinations. My gratitude is extended to my thesis chair, Dr. Angelos Barmpoutis, for his immense help in familiarizing me with the K3D software, and helping m e develop confidence in using digital technology. I am also thankful to Dr. Ben DeVane whose initial insight helped me hone my passions and apply them to my coursework and project. To Michael Baksh I'm truly appreciative of the time you set aside to show me how to use the Artec Scanner. Dr. James Oliverio and Tesa Brown both made sure that I acquired financial support to present my ideas and research at conferences, and for that, I'm truly thankful.


5 TABLE OF CONTENTS ACKNOWLEDGEMENTS ................................ ................................ ................................ ............. 4 LIST OF GRAPHS ................................ ................................ ................................ ........................ 6 LIST OF FIGURES ................................ ................................ ................................ ....................... 7 LIST OF TABLES ................................ ................................ ................................ ......................... 8 ABSTRACT ................................ ................................ ................................ ................................ ... 9 BACKGROUND ................................ ................................ ................................ .......................... 11 The Early (Mis)use of Anthropometry ................................ ................................ ..................... 11 Current Ant hropometry in Measuring Body Size and Shape ................................ .................. 13 Digital Technology and Indirect Anthropometric Methods ................................ ...................... 14 Indirect Anthropometry Using Handheld Technology ................................ ............................. 18 METHODOLOGY ................................ ................................ ................................ ....................... 19 STAGE I: Participant Recruitment and Ethnographic Interviews ................................ ............ 19 STAGE II: Traditional Direct Anthropometry Measures ................................ .......................... 22 STAGE III: Indirect Anthropometric Measurements ................................ ................................ 23 RESULTS ................................ ................................ ................................ ................................ ... 25 Direct Measurements ................................ ................................ ................................ .............. 25 Indirect Measurements ................................ ................................ ................................ ........... 27 Comparing Direct and Indirect Measures ................................ ................................ ............... 29 Determ ining Body Shape Using Indirect and Direct Anthropometry ................................ ....... 33 Indirect Anthropometric Rapid Assessment of Abdominal Obesity ................................ ........ 36 Participant Anthropometric Methodological Preferences ................................ ........................ 38 LIMITATIONS ................................ ................................ ................................ ............................. 40 DISCUSSION ................................ ................................ ................................ ............................. 42 BIBLIOGRAPHY ................................ ................................ ................................ ......................... 45 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ......... 49


6 LIST OF GRAPHS Graph 1 Direct and Indirect Methodological Trends and Discrepancies.. .......... .29 Graph 2 Body Size/Shape Predictors for Weight Related Disease Risk (BMI)... ...........34 Graph 3 Body Size/Shape Predictors for Weight Related Disease Risk (%BF)... .............34 Graph 4 Body Size/Shape Predictors for Weight Related Disease Risk (WHtR) ...... .3 4


7 LIST OF FIGURES Figure 1 Major Landmarks and Sites for Direct Anthropometry.....22 Figure 2 Avatar Wireframe View..24 Figure 3 Approximating Calf Girth .......28 Figure 4 Morphological Nuances Found During Direct Anthropometric Examinations...31 Figure 5 Indirect and Direct Posterior Comparison .......... 32 F igure 6 Body Shape Chart for Indirect and Direct Measures ....34 Figure 7 Avatar Difference Mapping across Body Shape .......37


8 LIST OF TABLES Table 1 Examinee Demographics ......20 Table 2 Direct Anthropometric Measures Sites 1 5 .26 Table 3 Direct Anthropometric Measures Sites 6 11 .................26 Table 4 Direct Anthropometric Measures Sites 12 17 ....26 Table 5 Indirect Anthropometric Measures ...27 Table 6 Disease Risk across Body Size and Shape ..............33 Table 7 Examinee Preferred Methods ...39


9 Summary of Project in Lieu of Thesis Presented to the College of Fine Arts of the University of Florida In Partial Fulfillment of the Requirements for the Degree of Master of Arts THE EFFICACY OF TRADITIONAL AND DIGITALLY DERIVED ANTHROPOMETRY AMONG BLACK WOMEN By Brittany Osbourne May 2013 Chair: Angelos Barmpoutis Member: Ben DeVane Major: Digital Arts and Sciences This project tests the efficacy of handheld technology against traditional direct methods to inexpensively gather anthropometric data from African American and Afro Caribbean women living in Florida, and to help create an ethnic specific digital database for future anthropometric survey use. Results demonstrate that indirect anthropometry that uses handheld technology has a shorter examination period, but provides less precision. This proposed project also explores the benefits and drawbacks of replacing 2D body image charts in clinical research and practice with charts derived from anthropometr y that uses handheld 3D technology. How participants interact with and experience handheld technology while undergoing anthropometric examinations are also r eported Based on methodological validity, direct anthropometry is more effective. More particip ants also prefer the direct anthropometric exam as opposed to the indirect methods derived from digital technology. However, although in its prototypical phase,


10 the technology used during this project presents unique opportunities t o compare individual bod y shape and size, which may serve as a powerful medical screening tool in the future.


11 BACKGROUND The Early (Mis)use of Anthropometry The develo pment of efficacious methods to measure the human body influe nces how we conceptualize physical health and disease risk and create human compatible technologies in contemporary society. Such measurements of the human body, skull, and face are collectively referred to as anthropometry ( Relethford 2010: 364 ). In the United States, anthropometry was central to the historical advancement of the sociopolitical and scientific belief held by early anthropologists and eugenicists, that human behavior was contingent on biologically predetermined race rather than the social environ ment (Ulijasjek and Komlos 2009, Baker 1998 ). Racial taxonomies were initia lly introduced by Carl Linnaus, and according to the article, "From a History of Anthropometry to Anthropometric History," anthropometry was the locus of this endeavor because it h elped to form, "the basis of scientific racism and eugenics in the late nineteenth century" ( Ulijasjek and Komlos 2009: 185). However, in the second half of the twentieth century, the Boasian anthropological school among others, challenged the S ocial Dar winist notion that behavior was determined by biological race rather than the social and material world. Anthropometric surveys were conducted by twentieth century anthropologists such as Franz Boas and Melville Herskovits, to prove that race was not biol ogically derived but socially constructed, and that the environment in the case of the US an environment that privileged some groups while disenfranchising others determined behavior and altered phenotype (Gershenhorn 2004 : 29). W idely known as the fathe r of American a nthropology, Franz Boas would conduct craniofacial and full body anthropometric measurem ents on the Shoshonean


12 people in North America and would later train other anthropologists such as Abram Harris, Zora Neale Hurston, Louis King, Melvill e Herskovits, and Greene Maxwell to perform similar anthropometric measures on European immigrants, African Americans, and other ethnic groups in the United States in the early twentieth century (Ulijasjek and Komlos 2009, Gershenhorn 2004 Boas 1899, Hers kovits 1932 ). These early twentieth century anthropometric surveys initiated by Boas would help debunk the racial ideology surrounding human typology, and facilitate a greater understanding for how morphological plasticity enables humans to adapt to their environments (Ulijasjek and Komlos 2009: 185). This early anthropometric data would also help pave the way for contemporary research that studies how racially hegemonic environments that socia lly include and exclude others w ould lead to a divided populace where some physically embody social privilege (i.e. healthy bodies) while others embody social inequality (i.e. bodies afflicted by health disparities) ( Farmer 2001 Krieger 1999 Gravlee 2009 Gulbas 2012 ). Contemporary research also draws upon The Human Adaptability section of the International Biological Programme (HAIBP) of the 1960s which employ ed an "explanatory framework" to further the understanding of human adaptation using anthropometric data; in particular, HAIBP studies conducted in "ninety thr ee nations between 1964 and 1974" would be used to chart and recreate human growth patterns across global populations over long periods of time (Ulijasjek and Komlos 2009 185 6, Collins and Weiner 1977, and Eveleth and Tanner 1976). The HAIBP's Worldwide V ariation in Human Growth of 1976 and 1990 which primarily used weight, height, and skinfold thickness, would serve as a foundational anthropometric publication that would influence the use of anthropometry in today's public health, epidemiology,


13 medical a nthropology, and other health social science fields (Ulijasjek and Komlos 2009: 187). It also would reinforce the current standpoint that variation in human growth patterns and morphology is not the result of intrinsi c racial difference, but rather, stems from the need for humans to increase their survival fitness through physical adaptation to a constantly changing social and natural environment. Consequently, since the environment continues to change either through global migration, geological change, tec hnological innovation etc. anthropometry remains a vital tool to document how humans effectively or ineffectively respond to that change. Current Anthropometry in Measuring Body Size and Shape D espite its early racial beginnings, a nthropometry has bec ome the most universal method to non invasively and inexpensively assess the size, proportion, and composition of the human body (WHO 1995). Today, a qualified anthropomet rist will employ several methodological measures height, weight, fat skinfolds, bone breadths, and girths among others to calculate the fat free mass (F FM ), such as muscle, bone, and fluid and fat composition of the body ( Macfarlene 2007, ISAK 2001 ). Within medical anthropology, as well as neighboring fields such as public health and medi cine, anthropometric data continues to elicit information about nutrition, exercise inactivity, fat distribution, and disease risk in individuals and populations ( Hussa in 2011) Two of the most common anthropometric measurements, height and weight, continu e to serve as the primary measures for calculating the Body Mass Index (BMI), which is considered "the most well known indicator of body fatness" (NHANES 1 1 ). However, just as environments change, the anthropometric methods used to measure physical respon se to that change must also adapt. Although the BMI continues to serve as a universal


14 index for weight classification, it has several limitations. By measuring total body mass, the BMI is unable to distinguish between FFM and body fat (Daniels 2009, Burkha user and Cawley 2008). Moreover, it does not distinguish between abdominal vi sceral fat and total adiposity; such a distinction is essential, because visceral fat, research suggests, is what directly increases a person's cardiovascular risk (Hamdy et al 20 06). Anthropometric measures that help to shape the BMI are not only used to classify body weight, but they are also used to help categorize body shape an d fat distribution. The Ashwell Shape Chart, among other bo dy image charts, rely on anthropometric measures such as height and waist circumference (WC) to calculate a waist to height ratio (WHtR) (Ashwell 2011). The Ashwell Shape chart distinguish es between four classic body types (1) banana (or chili pepper shaped), (2) pear, (3) pea r apple (or boxed shape), and (4) apple and serve s as a screening tool for clinicians and their patients to rapidly assess their central obesity ri sk (Ashwell 2011: 87). Although a useful tool body shape charts alone are poor proxies for anthropometric me asures that directly measure the various types of FFM, and distinguish between subcutaneous and visceral fat. Moreover, i n order for anthropometry to truly adapt in a way that effectively measures body composition, it had to draw upon a sociocultural environment that was increasingly relying on technology. Digital Technology and Indirect Anthropometric Methods Much of the measurements explained in the earlier sections are referred to as direct anthropometry This form of anthropometry has become sta ndardized by the International Society for the Advancement of Kinanthropometry (ISAK) and is considered to be the traditional standard for collecting anthropometric data (ISAK 2001).


15 With this method, measurements are obtained in areas of the body that ar e obscured by hair, and anthropometrists are able to directly assess depth of certain phenotypic attributes (Farkas and Deutsch 1996). However, direct methods are also associated with prolonged examinations, examiner skill variation and inconsistency, and unpredictable examinee behavioral response to measurements; such variables often dissuade clinicians and health researchers from using direct anthropometric methods for screening which contributes to a great er dependence on body weight and body shape char ts, notwithstanding their glaring limitations (Farkas and Deutsch 1996: 2 Park et al. 2012 ). Moreover, such a methodological approach also requires training on live human subjects, which can elevate the discomfort and anxiety of both examiner and examinee (Fourie 2011: 127). In contrast, indirect anthropometry involves collecting anthropometric data indirectly from the examinee usin g some form of technology. Four ubiquitous indirect methods are photogrammetry, cephalometry, bioelectrical impedance analysis (BIA), and laser surface scans (Farkas and Deutsch 1996: 2). Photogrammetry is an anthropometric method derived from 2D images, although emergent 3D photogrammetry is now being used to collect ant hrop o metric data using 3D images (Wong et al 2008 Weingberg et al 2006). Cephalometry is another indirect method, specifically used to gather craniofacial measurements; this method is normally performed using radiographs (or x ray photographs), and is frequently employed by orthodontists and other clinici ans (Farkas and Deutsch 1996) BIA calculates body composition, particularly body fat, by "sending a low level of electric current through the body and measuring the impedance ( Z ) of conducting tissues" (Haroun et al 2010: 1253). 3D digital surface scans c an either be performed using stereo photogrammetr y,


16 which is digital 3D photogrammetry, or surface laser scanning (Fourie et al 2011). Also medical imaging equipment, such as t he dual x ray absorptiometry (D XA), computed axial tomography (CAT), and magneti c resonance imaging (MRI), are commonly used in indirect anthropometry (Karisson et al 2012) Like direct methods, indirect anthropometry has its advantages. This methodological approach requires a shorter examination time, is less dependent on the examine e's behavior, and is accomplished using a simple measuring technique (Farkas and Deutsch 1996: 2 Haroun et al 2010 ). Moreover, whereas children tend to become impatient when undergoing traditional direct examinations, the rapid assessment achieved via som e forms of indirect anthropometry become more effective among the youth and those subjects with disabilities (Weinberg et al 2006: 478). Indirect methods are also noninvasive, and their reliance on digital technology allows for images containing anthropome tric measures to be digitally archived. Most importantly, the use of digital technology is considered an effective way to obtain precise and reliable measures, whereas measures obtained by anthropometrists who practice direct methodology may be subject t o greater inconsistencies and measurement error (Wong et al 2008: 233 ). However, although there are advantages to using indirect anthropometry, there are also drawbacks. Its reliance on technology inhibits its use in regions where electricity is highly vol atile This becomes problematic for health researchers such as medical anthropologists who study diverse populations, some of which come from developing nations or highly rural areas that may suffer from periodic electrical outages. Moreover, some indirect methods, such as cephalometry, emit radiation that may be


17 deleterious to examinees, which make multiple screenings that need to be conducted infeasible (Wong et al 2008, Park et al 2006). Also, despite the fact that BIA is inexpensive, the standard prediction equations derived from this method have not been rigorously validated on non white ethnic groups (Haroun et al 2010: 1253). In addition, although laser scanning has great precision, image capture is time consuming ; this may not be an issue for a n anthropometrist working with skeletal human remains; however, living subjects, particularly children and the disabled, may find it difficult to maintain the same stance for an extended period of time (Wong et al 2008: 233). Also, some indirect anthropome tric methods particularly those that use medical imaging equipment, can be very expensive to implement ( Wong et al 2008). Furthermore, many indirect methods do not capture hair, which may be a n important physical landmark in body image research particula rly when hair grooming, texure and styling is central to overall body satisfaction Moreover, very few methods have utilized handheld technology and those that do are primar il y us ed to measure craniofacial morphology rather than the whole body (Park et al 2006) This too can prove limiting for health researchers who must have mobility in order to conduct their work in various sociocultural contexts. Since many indirect methods were not initially created to measure the full body, particularly body composi tion, The Body Benchmark Study was a global collaborative study launched in 2007 to investigate if three dimensional body scanners were more effective as screening tools than the BMI in biomedical weight assessment (Hussain 2011 ). In particular, the scanne rs would calculate an individual's total body volume and segmental volume, and this would help establish the Body Volume Index (BVI), which has "the potential to be used as a long term computer based anthropometric


18 measurement for health care" (Hussain 20 11: 10). The BVI was officially launched in 2010 as an improved anthropometric measurement for healthcare and obesity predictability; in particular it would use, "a 3D scanner as opposed to manual measurementto calculate risk factors associated with a pe rson's body shape and type, through analysis of weight and body fat distribution" ( T he BVI has emerged in respon se to a global technological culture and is designed to eventually replace the BMI. However, it too has a drawback. Although the BMI has glaring limitations, it is an anthropometric method that can be performed using either direct or indirect methods, which allows it to be performed anywhere, and under various environmental conditions In contrast, BVI scanner s are full bodied and incapable of being portable. Unfortunately this presents a dilemma for health researchers and clinicians who would like to integrate this more precise and accurat e index into their work. Even though it arose out of a global technical culture, it s very design makes it inaccessible to certain regions, countries, and demographics. Although it has come a far way from the racial anthropometry of the nineteenth century, if it supplants the BMI in its current state, the BVI will disenfranchise certain people from efficacious health care and obesity screen ings, while privileging others namely those in developed nations that have regular access to health services, and reliab le electricity that can sustain a BVI scanning facility. Indirect Anthropometry Using Handheld Technology Within health research and clinical practice, there is a need for more anthropometric studies that integrate handheld technology. Reason being, alt hough


19 technology is as a powerful tool in anthropometric research, most indirect innovations like the Body Volume Index requires expensive full bodied scanning technology. This thesis project aims to help address the lack of handheld technology in indire ct anthropometry used for body image and obesity research and in clinical practice. In particular, this project tests the efficacy of traditional direct methods and indirect methods using handheld technology on ten African American and Afro Caribbean women living in Florida. This demographic was purposively chosen, because obesity is prevalent among black women, and their ethnic specific body ideals are not fully captured by 2D shape charts ( Ogden et al 2006 ). The following section provides an overview of t he methodology used during this project. METHODOLOGY To test direct anthropometry against indirect methods using handheld technology, the study primarily took plac e in Central Florida among ten African American (6) and Afro Caribbean women (4) T he project was informed by a research design and methodology commonly used within the field of medical anthropology. The study is broken down into three major stages. STAGE I: Participant Recruitment and Ethnographic Interviews The first stage involved recruiting a purposive sample of black women of various body sizes and shapes from local churches, universities, family, friends, and online social networking sites (SNSs) (Table 1)


20 Purposive sampling, as opposed to random sampling, was a more appropr iate method for recruiting participants, because such samples are widely used in: "(1) pilot studies, (2) intensive case studies, (3) critical case studies, and (4) studies of hard to find populations" (Bernard 2011: 145). T his project is an intensive case qualitative study that is not making generalizations across populations, but rather is rigorously testing and comparing multiple anthropometric measures on each examinee, which makes this form of sampling methodologically appropriate. Moreover, the small sample size is appropriate, because anthropological research indicates that 10 20, and even a sample size as small as 6, can be enough to, "uncover and understand the core categories in any well defined cultural domain or study of lived experience" (Bernar d 2011: 154). It is important to note that this research project is not only relying on the raw anthropometric data to determine which method is most effective; rather, as the abovementioned section indicates any method used must take into consideration t he examinee's behavioral response to that method in order to gauge its effectiveness. This small


21 sample was appropriate, because it allowed the researcher to conduct not only multiple anthropometric examinations across methods but it allowed for in depth participant observation and interviews with examinees to capture their phenomenological experience during the examination The phenomenological component of the project is critical, because much of the research on indirect methodological efficacy is focuse d on ways to improve the technology and limit human interaction rather than improve the human experience with th e technology that is being used. During examinations, unstructured interviews were conducted to learn how examinees perceived their bodies and o verall health. P articipants were also asked to openly share with the researcher their feelings about the current examination they were undergoing that involved both indirect and direct anthropometry. Preceding the examination, a survey was administered to participants, which lasted between 15 20 minutes. Each woma n was asked to provid e demographical information, along with their current lifestyle and exercise habits, and any weight related medical conditions they may have (e.g. high blood pressure, diabetes, body image disorders, or any other h ealth conditions related to body shape and weight). This survey was administered to help socioculturally situate the anthropometric data, and to determine whether body typ e or size (i.e. apples and hourglass versus obese or normal weight) is considered a m ore accurate measure for disease risk predictability Responses to the survey and unstructured interview were partially recorded in an excel spreadsheet, and all files were assigned a code number and were password protected for security purposes.


22 STAGE II: Traditional Direct Anthropometry Measures Seventeen measures were collected from each examinee. These measurements included: (1) body mass (kg), (2) stretch stature (cm), (3) triceps (mm), (4) subscapular (mm), (5) biceps (mm), (6) iliac crest (mm), (7) supraspinale (mm), (8) abdominal (mm), (9) front thigh (mm), (10) medial calf (mm), (11) arm girth relaxed (cm), (12) arm girth flexed and tensed (cm), (13) waist girth (min.)(cm), (14) gluteal girth (max.)(cm), (15) calf girth (max.)(cm), (16) biepico ndylar humerus (cm), and (17) biepicondylar femur (cm) ( Figure 1) This collection of direct measures was broken down into 2 basic measures ( weight and stature ) 8 skinfolds ( triceps, subscapular, biceps, iliac crest, supraspinale, abdominal,


23 front thigh and calf ), 2 bone breadths ( humerus and femur ), and 5 girths ( arm relaxed & flexed, waist, gluteal and calf ) (ISAK 2001). All sites were measured twice on one side of the body, excluding weight and stature, so as not to exceed 5% technical error of meas urement (TEM) (ISAK 2001, Nevill 2006). Since measurements were taken twice, less stature and body mass, direct measures totaled 32 0. The percentage of body fat (%BF) was then calculated using the ISAK recommended Yuhasz Method, which is % BF = (. 1548 x sum of all skinfolds (minu s biceps and iliac crest) + 3.580 ) (Yuhasz 1974). This %BF derived from direct methods was then compared to each participant's BMI and waist to height ratio (WHtR) or waist girth divided by stature Body size and body shape ca tegories were derived from each measure, and any discrepancies were recorded ( see Table 6 ). Slim Guide skinfold calipers were used to measure skinfolds, a Lufkin measuring tape was used for stature, a home scale measured body mass Lufkin W606PM tape and M yoTapes measured girths, and a small sliding cal iper was used for bone breadths; all equipment was properly calibrated prior to examinations. D irect anthropometric data was collected using the standardization methods created by the International Society f or the Advancement of Kinanthropometry (ISAK 2001). The research er underwent ISAK anthropometric training in the Summer of 2012. All participants were asked to disrobe down to their undergarments; however, sports bras and shorts were also permitted. All me asurements for each participant were recorded using an excel spreadsheet. STAGE III: Indirect Anthropometric Measurements After participants have been measured using direct anthropometry their bodies were scann ed using a Microsoft Kinect SDK. Microsoft Kinect is an active depth sensing


24 camera, which allows 3D body shape reconstructions to be obtained, "by combining multiple monocular views of a person moving in front of the sensor" (Weiss et al 2013: 99). To conduct indirect anthropometry usin g the Microsoft Kinect, a customized Kinesthetic Three Dimensional toolset ( K3D ) was used to generate 3D avatars for each participant (Barmpoutis 2013) Body scans were digitally archived and accompanied measurements were later exported into Microsoft Exc el To perform the body scan, participants were asked to stand in front of the Microsoft Kinect until they saw their 3D image from the neck down on the computer screen. To activate the camera, participants raised their left hands. Participants moved their bodies slightly in front of the camera to fill out their avatars, and they then raised their right hands to deactivate the Kinect Each avatar was saved in a folder as a .wrl file. The shape of each avatar was captured in wireframe mode, and a screenshot w as taken of each examinee (Figure 2).


25 The perimeter and volume for each body scan was saved as .txt file s For each .txt file nine measuremen ts were automatically generated: (1) torso, (2) left arm, (3) left forearm, (4) right arm, (5) right forearm, ( 6) left thigh, (7) left leg, (8) right thigh, and (9) right leg. This totaled 18 indirect measurement groups for each participant. All perimeter measures were in meters, and all volume measurements were in liters. Although ten women were scanned, two wishe d to withdraw their scanned images, and only have their direct measures included in the study Consequently, a total of ten women were measured using direct methods, and eight of their indirect measures are analyzed in this study. RESULTS Direct Measurements On average each examination took thirty five minutes to complete. In particular the marking of sites took approximately six to ten minutes; the first examination took roughly twelve minutes, and then the second examination took an additiona l ten to twelve minutes. The amount of time it took to complete a direct anthropometric exam correctly was very time consuming comparable to the indirect method that followed. This time constraint became even more pronounced among older participants who s uffered from arthritic pain. The following tables provide the average of the two examinations.


26 Although the direct method was very time consuming, it produced information that the indirect method using the Kinect simply could not do. For example, to measure bone breadth, an anthropometrist must feel for certain indentations between and or on bone before positioning the sliding caliper. Moreover, skinfolds require that the


27 examiner pinch the skin and subcutaneous fat using a caliper. 3D t echnology t hat only scans the surface of the body can not perform this. Consequently, such human to human contact is an essential prerequisite for an accurate bone breadth or skinfold measurement to be obtained. Moreover, although a second examination prolongs the ant hropometric screening, it increases the precision of measurements by accounting for the technical error of measurement. Indirect Measurements In contrast, whereas the direct methods took on average thirty five minutes to complete, once the Microsoft Ki nect was properly setup, the body scan took anywhere between a few seconds to five minutes. The reason why some indirect sessions took five minutes was because some participants were scanned as much as four or five times before a suitable avatar was genera ted The brief time duration associated with this method was not surprising, since much of the literature substantiates that indirect anthropometric methods have a shorter duration than direct anthropometric examinations. The following table provides the a ctual measurements derived from this method:


28 It is important to note that the .txt files only provide values for the front half of the body, because K3D only scans the front of the body. Multiplying the anterior values by two approximates values for the posterior side of the body. A fter this was done a ll volume values were derived by adding all of the numbers in the volume.txt file (for avatar volume), and the torso row (for torso volume). Girths used the perimeter.txt files. Within the .txt file, a number represents the perimeter of each edge loop and edge loops are arranged in rows from left to right; the left number represents the top of the body segment, and the number furthest to the right represents the number for the bottom of the segment. For example, for BE0306 the top of her left leg is the edge loop value furthest to the left in the .txt file (.22 m 2 = .44m ) and the bottom of her leg, near the ankle, is the value the furthest to the right (.15 m 2 = .30m ). An anatomical assessment of the human body confirms that the upper leg is wid er than the lower leg. All girth measurements were calculated using the corrected perimenter.txt files for the legs (for calf girth), arms (arm girth relaxed) and torso (for waist girth). Since calf girth is taken at the maximum width of the calf, the fol lowing equation was used: calf girth = (Right Leg Q1 )/2. Q1 represents the median of the first quadrant, or the midpoint (mp) of the upper calf. T he value s for edge loops are illustrated by a relative line and are not actually shown on the scanned image; i nstead they are exported as .txt files. O ne can either manually count each individual loop (which is not very


29 feasible) or approximate by exporting the .txt file into an excel spreadsheet and multiplying all values by two to represent the fully body, rather than the anterior side only. Consequently, t he Q1 of the right leg was the desired number, because the widest part of the calf is approximately located in the upper half of the leg (Figure 3). Arm girth (relaxed) was calculated by taking the averag e of both the right and left arm median. The median (or middle number) rather than Q1 was the preferred measure, because the direct measure of the arm girth is taken at the midpoint of the superior part of the acromion border, to the proximal and lateral p oint of the radial head (ISAK 2001). In other words, arm girths are taken at the midpoint of the arm. The waist girth is taken at the most narrow point of the waist, and the median edge loop of the torso was the approximate measure used. Comparing Direct and Indirect Measures I ndirect measures could not be generated for skinfolds or bone breadth s so waist, arm, and calf circumferences were the only measures that could be compared for each examinee (Graph 1)


30 There was a huge technical error of measu rement (TEM) between the indirect and direct methods used. As already mentioned, a good %TEM should be no greater than 5%; however, this general rule applies to checking the accuracy of the intra tester ; this is when one person performs both examinations. In this study's case, we take the average of the examiner's direct measurements and compare them to the indirect measures derived from the K3D body scans to calculate the inter tester %TEM or the error of measurement between exams performed by two separate examiners. A good inter tester %TEM is 7.5%. However, both 5% and 7.5% is the maximum TEM allowed for skin folds. Other anthropometrical measures, like girths, require an inter tester % TEM of 1.5% or less (ISAK 2001 Sicotte 2005 : 3 ). %TEM was calculated by (1) finding the deviation (d) between the indirect and direct measures for a site on each individual (2) taking t he square root of the deviation (d 2 ) (3) su m ming all squared deviations ( d 2 ) and apply ing them to this equation : ( d 2 / 2n), where (n) equals the total number of examinees measured ( Perini 2005, ISAK 2001). This results in the absolute TEM. To find the relative or %TEM, the variable average value (VAV or the mean ) is first calculated by finding the average of the indirect and direct measures for the given site on each woman. Add the measurement averages for all women (8 total), and divide by 8. This final equation is used to find %TEM: [(absolute TEM/VAV)*100] (Perini et al 2005, ISAK 2001). The calculated %TEM for calf girths, arm girths (relaxed), and waist girths was 7 .4 %, 17.7 %, and 7 .2 % respectively The corresponding absolute TEMs are 2.9 cm, 5.4 cm, and 6.4 cm. Based on these results, there was a greater relative technical er ror of measurement found among arm girths (relaxed), and waist girths had the least %TEM.


31 This major %TEM may signify that there are shortcoming s in either one or both methods Such a drastic discrepancy can either result from the examiner 's inaccuracy during dir ect anthropometric exams, or the result of the Kinect and K3D software not fully or accurately capturing the examinee's body during the surface scan. As already mentioned the Kinect using the K3D toolset only captures the front of t he person, while digitally approximating the posterior. Perimeter and volume files derived only account for the front of the scanned body, and are doubled to achieve an approximate 3D representation. However, this half capture approach the K3D uses does no t take into consideration that the front of the body is morphologically different than the back. This is why anthropometrists who use direct methods, when measuring the arm for example, take skin fold measures at the tricep and bicep sites. The half captur e approach also ignores morphological nuances that play a major role in accu rately measuring girth For example, when taking the direct measure of waist girth, a trained anthropometrist will be cognizant of the natural sloping of the back ( Figure 4)


32 Th is half capture approach also proves to be problematic, because body scans do not reflect body shape variation of examinees, particularly those who are well endowed in the gluteofemoral region of their bodies. Using this technique, women with well endowed posteriors will intrinsically not have an accurate 3D represent ation of their bodies (Figure 5 ). H owever, the girths compared in G raph 1 do provide a unique opportunity to assess trends between methods, and point out possible outlier direct or indirect measurements within methods that would need to be measured again for increased validity. This would prove useful for anthropometrists who recognize the limitations of the current K3D software, but would like to perform a visual assessment of th ose measurements they knew were less than precise during examinations.


33 Determining Body Shape Using Indirect and Direct Anthropometry Part of this research project aimed to evaluate which method would better inform body image charts and predict disease risk To do this, the BMI, WHtR, and %BF were calculated from direct measures (Table 6). As already noted, the body mass index is not an accurate estimate for body shape or fat because it only determines total body mass and does not distinguish between t otal adiposity and fat free mass The percentage of body fat (%BF) could not be used for comparison purposes, because skinfolds could not be obtained using indirect anthropometric measures. The WHtR was the only measure that could be used, because waist gi rths could be obtained using both methods. Graphs 2 4 illustrate how these predictors determine body shape and disease risk derived from direct measures. To compare WHtRs across methods, the direct measure for stature was used for both the body scan and ac tual examinee, because avatar he ight does not include the head (Figure 6 ).




35 As Table 6 shows, eight participants, including some who were obese, did not suffer from any weight related medical conditions. Most women self reported that they exercised regularly and led healthy lifestyles which may indicate that their body shape/size and associated disease risk is predicated on genetics or some other extrinsic sociocultural factors. Based on Table 6, most women were classified as having an hou rglass shape, including those who may be considered overweight by BMI standards. In contrast, women with pear shapes were generally considered to be either seriously overweight or obese. However, when viewing Figure 6 it can be noted that the ir WHtR score s from direct measuring fell below the median value associated with their height, which according to the Ashwell Shape Chart, would place them in the pear apple range (Ashwell 87: 2011). There was an obesity classification consensus found across all predic tors for women with apple shapes. Moreover, although WHtR was the only predictor compared across methods, there was a strong correlation between it and BMI and %BF, particularly for obese women. W hen compared across methods, there was 50% consensus in WHt R ra t ios and body shape classifications. Of those women who were scanned, direct anthropometry classified four of them as hourglasses, but only three were classified as such using indirect methods. Also, direct measures categorized only two women as pears, whereas four were classified as pears with indirect methods. Finally whereas direct measures classifed only two scanned women as apples, only one is classified as an apple using indirect measures. In sum, four women (EH0124, MB0814, DH0218, and MN1129) ha d the same body shape classifications across methods. However, the other four (VO0720, YG1026, BE0306, and RV0212) had body shape discrepancies across


36 anthropometric methods. Althogh VO0720's discrepancy was quite significant, both methods placed here with in the hourglass pear range, which is considered a healthy body type. However, RV0212, has a pear apply body type with direct measures, but a very distinct pear shape using indirect anthropometry. In contrast, YG1026 is classified as a medial pear using di rect methods, and a marginal apple with indirect methods. Such discrepancies could be contingent on the high %TEM between both methods, or it could result from having a body shape chart that narrowly classifies the female body, by viewing body shape in mut ually exclusive, rather than overlapping and highly diverse categories. Indirect Anthropometric Rapid Assessment of Abdominal Obesity The K3D software includes a useful difference mapping tool that allows avatars to be compared to each other. This enabl es anthropometrists to metaphorically compare apples to oranges, or in this case, women with diverse body shapes. Much of the body image research indicates that apple body types generally deposit fat in their abdomens, which can lead to central (or abdomin al) obesity and increased risk of metabolic syndrome ( Shearer 2013 ). In contrast women with pear shapes tend to accumulate fat deposits in the gluteofemoral region (e.g. the buttocks, hips, and thighs); research indicates that the latter fat distribution is indicative of greater fertility, fitness, and overall healthy ( Cashda 2008 ). The hourglass is the only body shape however, that tends to ha ve an equal distribution of fat Below are the results from the difference maps that compare apple, pear, and hou rglass body types.


37 The first two avatars illustrate that indeed women classified as having apple bodies have more central fat deposition than women with hourglass body types. Keep in mind, that RV0212 and BE0306 had body shape discrepancies across me thods. However, by using the difference mapping there may be a stronger argument made for direct anthropometry, which classified both women as apple shaped. In contrast, the last avatar indicates that RV0212 has less abdominal adipose f at than pear shaped YG1026 This may be an indication that indirect methods were more precise with these measures. However, when viewing Table 5, YG1026 is classified as either being seriously overweight or obese, and RV0212 is classified as being obese across predictors. In Figure 6 there is a distinct classification shift for YG1026, whereas the discrepancy in RV0212 is smaller and takes place at the intersection of pear and apple classifications. It is important to note that when RV0212 is compared with a true hourglass, s he is considered an apple; however, when she is compared with someone


38 who also has a body shape discrepancy, she becomes a pear. Such discrepancies could have resulted from posterior less body scans, or examiner inconsistency during direct screening. As pr eviously mentioned, the body shape chart itself could have intrinsic flaws, or WHtR may not be a good tool for comparing women who are either seriously overweight or obese. There are many reasons for why this discrepancy may exist. Notwithstanding, these three difference maps are a great way to assess and compare body shape and size. If the K3D software was further tweaked, it could capture more precise full bodied scans. This would allow women of various body weights who were directly measured and found t o be true pear, hourglass, or apple body types to become standardized avatars in clinical screenings. Clinicians trained in these indirect scanning methods could rapidly assess what body type range their patient falls within, and how close or far they devi ate from the ideal range of fat distribution. Participant Anthropometric Method ological Preference s Based on the current indirect and direct anthropometric methods used, traditional direct anthropomet ry is a more effective tool than indirect anthropom etry that us es handheld digital technology. Methodologically, this is primarily due to the fact that the K3D software only captures the anterior angle of the body, while digitally reconstructing the posterior. This skews findings, and make any measurements derived from such body scans questionable. However, in addition to method precision, it i s important to understand the phenomenological experience of participants during anthropometric examinations in order to determine which method they preferred (Table 7)


39 As Table 7 shows, 60% of participants preferred the traditional direct method. 30% preferred the indirect method, and 10% did not care. It is important to note that 2 of those who preferred the i ndirect method were younger than most women in the study. Although research purports that younger children are impatient with direct anthropometry, young adults may share this feeling. Those who preferred the direct method, however, crosscut all ages. Some women preferred the latter meth od because they believed it was a more accurate way of measuring the body. The oldest of the participants, BE0306, was not perturbed by the long duration, because she used the time to engage in conversation with the examiner. Lack of privacy was another re ason for why some women preferred one method over the other. CM0703, for example, was uncomfortable with her body scan being archived indefinitely by the examiner. Even if it was a headless 3D representation of her actual self, she still claimed ownership of the


40 image as her own body. In contrast, RV0212 felt like the direct examination intruded upon her personal space. The pulling and pinching of her body made her metaware of her "fat which made her extremely uncomfortable. Many of the women who chose direct anthropometry as their preferred method did so because they were dissatisfied with the way they were portrayed using 3D technology. Some women felt that their body scans were not accurate representations of who they really were. MB0814 felt that the use of beige hands were not racially ambiguous, but showed preference for individuals with lighter skin. Similarly VO0720 and RV0212 believed that their body scans stripped them of their curves and femininity. Fin ally, SS1111 compared her body scan exper ience to having a picture taken and uploaded to a social networking site (SNS). She wanted to share authority with the examiner in determining which scan, if any, is shown. LIMITATIONS Initially, this project intended to use the Artec Studio scanner, wh ich has greater precision. However, the scanner was appropriate for only medium objects, which would have worked well for specific body areas, rather than ful l body scans. Due to this limitation the K inect scanner was used as a substitute. However, the pe r pixel depth resolution of the Kinect camera is one centimeter when the camera is positioned approximately 2 meters away ( Wilson 2010 ). Moreover, the depth precision of the camera decreases as the examinee's distanc e from the camera increases. This limita tion in the camera's depth precision may also play a role in the low level of precision of the indirect methods used.


41 During the conceptual phase of this project, body scans were going to be exported as .obj file s to Maya, a 3D modeling software. The purp ose of this exportation was to indirectly measure avatars using 3D modeling tools, and "tweak" them to have a more accurate representation of the participant. However, the researcher opted instead to use the K3D software's built in measurement tools that automatically assess the perimeter and volumes of each avatar; however, such actual measurements were only viewable by .txt file export and were only represented as relativ e lines within the actual software. Moreover, Maya was not used to "tweak" avatars, because such remodeling would be contingent on the researcher's artistic discretion and expertise. Remodeling avatars would have only resulted in a realistic vision, not re plica, of actual participants and it would not have allowed for a direct comparison between indirect and direct methods. The K3D software and Microsoft Kinect camera used to compare, measure, and archive the digital scans had several limitations. As alrea dy mentioned, the posterior side of participants could not be captured using this technology. Rather the examiner had to rely on the software 's reconstruction of these areas. Consequently, most women with well endowed buttocks had flatter posteriors in th e accompanying scan. Moreover, the use of beige hands made the examination experience uncomfortable for some women Many of the women who agreed to participate in the study had skin color that ranged in various shades of brown, and the lack of race neutral colors to them was an indication that th e technology had inherent bias As far as direct anthropometric methods, examinations were very time consuming, and some of the equipment used, although ISAK approved, was not the


42 most ideal instruments for meas urement. For example, although ISAK has designated Slim Guide calipers as a tool to accurately measure skinfolds, the gold standard for skinfold measuring is the Harpenden caliper (ISAK 2001) This caliper was not used because it was too expensive and was outside of the examiner's budge t. Although a trained anthropometrist can serve as both examiner and data recorder having two anthropometrists, one examiner and one recorder, helps to increase precision and validity. Reason being the recorder's role is, to assist' the examiner in obtaining correct measurements" ( NHANES 2007: 1 4 ). The recorder not only records the actual measurements as they are taken, but he or she helps position instruments on the examinee, and quickly identifies errors in measurement at sites that need to be measured again. In the case of this study, the researcher acted as both examiner and recorder. Moreover although this project did not need a large sample size, more participants could have enriched the data collected. In future re search, a larger sample could be used to make generalizations, some thing this study could not do. DISCUSSION From its early use within the United States, anthropometry has been both praised and contested as a tool for measuring body composition. In anthropometry's developmental phase anthropologists both thwarted and furthered the ideological belief that race was biological rather than constructed by one's own material and social environment. However, more anthropometric research reinforces the reality that race is a social construct and our physical difference is more indicative of our body's re sponse to envi ronmental change. Physical difference is not only found between socially constructed races or ethnicities, but as this study has shown, it is found within them.


43 Morphological variation does not only happen on a population level, but it occurs between indi viduals. Such individual difference provides the opportunity to understand how different people experience and interact with different anthropometric methods. For example, h ow efficacious are direct or indirect anthropometric methods that are precise, but leave individuals feeling invisible, unfeminine, or "fat"? What criterion do we as anthropometrists use to determine methodological efficacy the method, the person, or both? Certainly, methods can be determined to be effective based on precision and valid ity, but if integrated into clinical settings, will it cultivate or exacerbate the clinician patient relationship? Will patients be mor e or less compliant with follow up examinations or receptive to advice, given their initial experience? Perhaps some indi viduals like YG1026 are most concerned with their examination's accuracy, which could provide them with deeper insight into their bodily health. Indeed, g reater precision leads to a more accurate virtual representation of the body. This reality resonated w ith most participants, and is why direct anthropometry is considered their preferred form of examination, and is considered a more effective tool of measurement. Consequently, this study argues that both the person and technology are fundamental to assessi ng the efficacy of direct and indirect anthropometric methods. Even though the indirect method used in this study was not as effective as the direct anthropometric methods due to its half capture approach there is room for improvement and future applicat ion. Capturing a full 3D representation of the body would allow body scans to be measured against measures de rived from direct anthropometry, because it would take into consideration morphological nuances and


44 diversity. It would also make the difference ma pping feature in the K3D software a powerful tool in clinical practice research and education Anthropometric exam simulations could be conducted on interactive 3D avatars as opposed to real humans. Students within the health field who use anthropometry could identify major landmark sites, and perform various direct measures indirectly. If avatars can be animated, students could also be tested on their cultural competency and interaction with the examinee. Furthermore, proponents of 3D body image charts that assess body shape, size, and disease risk would have a stronger platform to argue against 2D charts that fail to capture fat depositions in healthy areas (i.e. the gluteofemoral region). If the K3D software can capture the posterior angle of individu als, and take into consideration the individual's phenomenological experience, it could become a very powerful anthropometric screening tool that could be just as effective, if not more so, than direct anthropometry.


45 BIBLIOGRAPHY Ashwell, Margaret. 2011 "Shape: The Waist to Height Ratio is a Good, Simple Screening Tool for Cardiometabolic Risk." Nutrition Today 46(2): 85 89. Baker, Lee. 1998 From Savage to Negro Berkeley: University of California Press. Barmpoutis, Angelos. 2013 "Automated Human Avatar Synthesis for Obesity Control using Low Cost Depth Cameras." Stud. Health Technol. Inform. 184: 36 42. Bernard, Russell H. 2011 Research Methods in Anthropology: Qualitative and Quantitative Approaches 5 th Edition Lanham: AltaMira Press. Boas, Franz. 1899 "Anthropometry of Shoshonean Tribes." American Anthropologist, New Series 1(4): 751 758. 2013 "Body Volume Index." website. Burkhauser, Richard and John Cawley. 2008 "Beyond BMI: The value of more accurate measures of fatness and obesity in social science research." Journal of Health Economics 27: 519 529. Cashda, Elizabeth. 2008 "Waist to Hip Ratio across Cultures: Trade Offs between Androgen and Estrogen Dependent Traits." Current Anthropology 49(6): 1099 1107. Collins, K.J. and J.S. Weiner. 1977 Human Adaptability: A History and Compendium of Research London: Taylor and Francis. Daniels, Stephen. 2009 "The Use of BMI in the Clinical Setting ." Pediatrics 124(1): S35 S41. Eveleth, P.B. and J.M. Tanner. 1990 Worldwide Variation in Human Growth. 2 nd Ed Cambridge: Cambridge University Press. Farkas, Leslie and Curtis Deutsch. 1996 "Anthropometric Determination of Craniofacial Morphology."


46 American Journal of Medical Genetics 65:1 4. Farmer, Paul. 2001 Infections and Inequalities: The Modern Plagues Berkeley: University of California Press. Fourie, Zacharias et al. 2011 "Evaluation of anthropometric accuracy and reliability using diffe rent three dimensional scanning systems." Forensic Science International 207(1 3): 127 134. Gershenhorn, Jerry. 2004 Melville J Herskovits: And the Racial Politics of Knowledge. Lincoln: U of Nebraska Press. Gravlee, Clarence. 2009 "How Race Becomes Biology: Embodiment of Social Inequality ." American Journal of Physical Anthropology 139: 47 57. Gulbas, Lauren. 2012 "Embodying Racism: Race, Rhinoplasty, and Self Esteem in Venezuela." Qualitative Health Research 1 10. Hamdy, Osama, Sriurai Porramatikul, and Ebaa Al Ozairi. 2006 "Metabolic Obesity: The Paradox Between Visceral and Subcutaneous Fat." Current Diabetes Review 2(4): 1 7. Haroun, Dalia et al. 2010 "Validation of Bioelectrical Impedance Analysis in Adolescents Across Differe nt Ethnic Groups." Obesity 18(6): 1252 1259. Herskovits, Melville. 1932 "Negro Year Book: An Annual Encyclopedia of the Negro, 1931 1932." American Anthropologist, New Series 34(3): 528 530. Hussain, Tahir. 2011 Anthropometry in Physical Education and Sports New Delhi: Sports Publication. International Society for the Advancement of Kinanthropometry (ISAK). 2001 International Standards for Anthropometric Assessment Australia: ISAK Karisson, Ann Katrine et al. 2012 "Measurements of total and r egional body composition in preschool children: a comparison of MRI, DXA, and anthropometric data." Obesity Accepted Article: 1 26.


47 Krieger, Nancy. 1999 "Embodying Inequality: A Review of Concepts, Measures, and Methods for Studying Health Consequenc es of Discrimination." International Journal of Health Services 29(2): 295 352. Macfarlane, Duncan. 2007 "Can bioelectrical impedance monitors be used to accurately estimate body fat in Chinese adults?" Asia Pac J Clin Nutr 16(1): 66 73. National Hea lth and Nutrition Examination Survey (NHANES) 2007 "Anthropometry Procedures Manual." Center for Disease Control. Nevill, Alan, Arthur Steward, Tim Olds, and Roger Holder. 2005 "Relationship Between Adiposity and Body Size Reveal Limitations of BMI." American Journal of Physical Anthropology 129: 151 156. Ogden, Cynthia et al. 2006 "Prevalence of Overweight and Obesity in the United States, 1999 2004." The Journal of the American Medical Association 295(13): 1549 1555. Park, Hyun Jae. 2012 "Correl ation between Direct Anthropometry and Di3D Camera System." Journal of Hard Tissue Biology 21(1): 87 92. Park, Hee Kyung et al. 2006 "Use of hand held laser scanning in the assessment of craniometry." Forensic Science International 160: 200 206. Perin i, Talita et al 2005 "Technical error of measurement in anthropometry." Rev Bras Med Esporte 11(1): 81 85. Relethford, John. 2010 The Human Species: An Introduction to Biological Anthropology. New York: McGraw Hill. Shearer, E.S. 2013 "Obesity anesthesia: the angers of being an apple." British Journal of Anesthesia 110(2): 172 174. Sicotte, Maryline et al. 2010 "Reliability of anthropometric measures in a longitudinal cohort of patients initiating ART in West Africa." Medical Research Methodology 10(102): 1 9.


48 Ulijaszek, Stanley and John Komlos. 2010 "From a History of Anthropometry to Anthropometric History." Human Variation: From the Laboratory to the Field CG Nicholas Mascie Taylor, Akira Yasukouchi, Stanley Ulijaszek Eds. Boca Rotan: CRC Press. Pgs 183 198. Weinberg et al. 2006 "Anthropometric Precision and Accuracy of Digital Three Dimensional Photogrammetry: Comparing the Genex and 3dMD Imaging Systems with One Another and with Direct Anthropometry." The Journal of Craniofacial Surgery 17(3): 477 483. Weiss, Alexander et al. 2013 "Home 3D Body Scans from a Single Kinect." Consumer Depth Cameras for Computer Vision A. Foassati et al. eds pgs 99 117. Wilson, Andrew. 2010 "Using a Depth Camera as a Touch Sens or." Devices and Algorithms 69 72. World Health Organization (WHO). 1995 Physical Status: The Use and Interpretation of Anthropometry. Report of a WHO Expert Committee Technical Report Series. Geneva: WHO. Wong et al. 2008 "Validity and Reliability of Craniofacial Anthropometric Measurement of 3D Digital Photogrammetric Images." Cleft Palate Craniofacial Journal 45(3): 232 239. Yuhasz, M.S. 1974 Physical Fitness Manual London and Ontario: University of West Ontario.


49 BIOGRAPHICAL SKETCH Brittany Osbourne is a PhD candidate specializing in medical anthropology at the University of Florida. She holds bachelor degrees in English and anthropology, and earned a Masters of Fine Arts in Creative Writing from the University of Central Florida. Her academic passions for anthropology and t he arts have led her to study abroad in Ghana and Camb ridge University, and present on her research at national academic and nonacademic conferences, as well as an international con ference at the University of Lisbon in Portugal. Since pursuing her PhD, Brittan y has conducted workshops on body image and undergone rigorous ISAK anthropometric tr aining at Princeton University. She has also served as a biometric screener in the Shands Employer S urvey. Currently, she works as the assistant web editor for, a site dedicated to social research methods funded by the National Science Foundation. When she finds time away from her studies, Brittany enjoys revising her MFA masters thesis novel, v/blogging about body image topics prevalent in the African Diaspora, participating in local and online community activism, spending quality time with her f iancÂŽ, family, and friends, and planning the day she jumps the broom with her future h usband!

xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EA5KIRGJW_K9863P INGEST_TIME 2016-04-21T19:43:52Z PACKAGE AA00017015_00001

xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd