Welcome to MorphoPASSE
The Morphological Pelvis and Skull Sex Estimation Database
Developing a free interactive morphological database for estimating sex in modern adults
PI: Dr. Alexandra R. Klales
Research Fellow: Stephanie J. Cole
Funding agency: National Institute of Justice (2015-DN-BX-K014)*
*This research was funded by National Institute of Justice Grant 2015-DN-BX-K014. Opinions or points of view expressed in this research represent a consensus of the authors and do not necessarily represent the official position or policies of the U.S. Department of Justice or the National Institute of Justice. Any products and manufacturers discussed are presented for informational purposes only and do not constitute product approval or endorsement by the U.S. Department of Justice or the National Institute of Justice.
The primary aim of this project is to examine secular changes, population variation, observer error, and the effects of asymmetry on sex classification using the eight most popular morphological traits of the skull (five used in the Walker 2008 method) and the pelvis (three used in the Klales et al. 2012 method).
With the data and results of these analyses, the secondary aim of this project is to develop a free, interactive morphological database, based on these standardized methods. With the program, practitioners can enter, compare, and analyze morphological traits of their unknown forensic cases to a large, modern sample with known demographic data in order to more accurately and more easily estimate sex in a manner compliant with Daubert. Practioners will be able to score the eight cranial and pelvic traits in their unknown individual and then use the temporal and population specific equations to predict sex membership with concordant posterior probabilities and accuracy rates.
(excerpt from NIJ grant proposal)
The primary goal asked of forensic anthropologists is the estimation of a person’s biological profile in order to help law enforcement determine the identity of unknown human remains encountered in forensic contexts. Sex estimation is an important component of the biological profile and many methods have been developed, tested, and utilized to estimate sex using different regions of the human skeleton. Methods of adult sex estimation using the pelvis, long bones, and cranium have been well accepted and utilized within forensic anthropology. Despite the availability of quantitative methods for sex estimation, many forensic anthropologists continue to rely on qualitative (morphological/non-metric) traits for sex estimation. In a survey of current American Academy of Forensic Science members, most practitioners (63.6%) prefer to use both qualitative and quantitative methods; however, when both are not used, qualitative methods (23.9%) were preferred nearly twice as often as metric methods (12.5%) (Klales 2013).
Benefits of morphological methods include ease of use, relatively quick application, no need for specialized equipment, and applicability with fragmentary remains. Unfortunately, many of the qualitative methods used for sex estimation are based on subjective interpretations of cranial and pelvic traits (Rogers and Saunders 1994; Williams and Rogers 2006). Because of this, attempts to standardize the use of morphological traits has resulted in the creation of methods that rely on standardized ordinal scoring and statistical methods of sex classification (e.g, Walker 2008; Klales et al. 2012), thereby making the methods compliant with the Daubert recommendations (Daubert vs. Merrell Dow Pharmaceuticals 1993). According to the Klales (2013) survey, the most preferred traits for morphological cranial sex estimation are the traits depicted in Buikstra and Ubelaker (1994) and subsequently utilized by Walker (2008). Likewise, survey participants indicated that the three traits originally described by Phenice (1969) and then modified by Klales et al. (2012) are the most popular for morphological pelvic sex estimation (Klales 2013). Only a few studies have validated or attempted to recalibrate the predictive sex estimation equations presented in Walker (2008) and Klales et al. (2012) and the results have widely varied (e.g., Roth et al. 2013; Stull et al. 2013; Garvin et al. 2014). Despite being popular and frequently applied to active forensic cases, secular change and population variation in the traits utilized in Walker (2008) and Klales et al. (2012) have not yet been fully explored. Furthermore, many of these morphological traits are bilateral and the effects, if any, of asymmetry have not been evaluated for their impact on sex classification accuracy.
Given the survey results mentioned above, it would appear that most forensic scientists are, in fact, using both types of methods, albeit perhaps with a preference for qualitative methods, for forensic case work. Quantitative sex estimation methods were consolidated into the computer program FORDISC in 1993 (Jantz and Ousley 2005). Cranial and postcranial metric data from 13 modern populations (8 male groups and 5 female groups) have been integrated into the interactive program (Jantz and Ousley 2005). Forensic scientists can enter the measurements of their unknown individual into the program and then compare their case to those individuals within the known reference samples for sex and ancestry classification using linear discriminant function analysis. Users are then provided with discriminant function classification accuracy, a posterior probability of group membership, and several typicalities which the practitioner can then interpret for combined sex and ancestry estimation. Within two short decades, FORDISC has become the number one method for metric assessment of sex and ancestry (Klales 2013). The ease of use and inclusive nature of the program is likely why FORDISC is the number one method for metric assessment of the sex and ancestry parameters of the biological profile (Klales 2013). Unfortunately, a similar program for morphological methods does not yet exist; therefore, forensic scientists must rely on the equations provided in the original Walker (2008) and Klales et al. (2012) publications and then must evaluate sample and statistical appropriateness when applying the method to their unknown individual....until now!
PI: Dr. Alexandra R. Klales
Research Fellow: Stephanie J. Cole
Funding agency: National Institute of Justice (2015-DN-BX-K014)*
*This research was funded by National Institute of Justice Grant 2015-DN-BX-K014. Opinions or points of view expressed in this research represent a consensus of the authors and do not necessarily represent the official position or policies of the U.S. Department of Justice or the National Institute of Justice. Any products and manufacturers discussed are presented for informational purposes only and do not constitute product approval or endorsement by the U.S. Department of Justice or the National Institute of Justice.
The primary aim of this project is to examine secular changes, population variation, observer error, and the effects of asymmetry on sex classification using the eight most popular morphological traits of the skull (five used in the Walker 2008 method) and the pelvis (three used in the Klales et al. 2012 method).
With the data and results of these analyses, the secondary aim of this project is to develop a free, interactive morphological database, based on these standardized methods. With the program, practitioners can enter, compare, and analyze morphological traits of their unknown forensic cases to a large, modern sample with known demographic data in order to more accurately and more easily estimate sex in a manner compliant with Daubert. Practioners will be able to score the eight cranial and pelvic traits in their unknown individual and then use the temporal and population specific equations to predict sex membership with concordant posterior probabilities and accuracy rates.
(excerpt from NIJ grant proposal)
The primary goal asked of forensic anthropologists is the estimation of a person’s biological profile in order to help law enforcement determine the identity of unknown human remains encountered in forensic contexts. Sex estimation is an important component of the biological profile and many methods have been developed, tested, and utilized to estimate sex using different regions of the human skeleton. Methods of adult sex estimation using the pelvis, long bones, and cranium have been well accepted and utilized within forensic anthropology. Despite the availability of quantitative methods for sex estimation, many forensic anthropologists continue to rely on qualitative (morphological/non-metric) traits for sex estimation. In a survey of current American Academy of Forensic Science members, most practitioners (63.6%) prefer to use both qualitative and quantitative methods; however, when both are not used, qualitative methods (23.9%) were preferred nearly twice as often as metric methods (12.5%) (Klales 2013).
Benefits of morphological methods include ease of use, relatively quick application, no need for specialized equipment, and applicability with fragmentary remains. Unfortunately, many of the qualitative methods used for sex estimation are based on subjective interpretations of cranial and pelvic traits (Rogers and Saunders 1994; Williams and Rogers 2006). Because of this, attempts to standardize the use of morphological traits has resulted in the creation of methods that rely on standardized ordinal scoring and statistical methods of sex classification (e.g, Walker 2008; Klales et al. 2012), thereby making the methods compliant with the Daubert recommendations (Daubert vs. Merrell Dow Pharmaceuticals 1993). According to the Klales (2013) survey, the most preferred traits for morphological cranial sex estimation are the traits depicted in Buikstra and Ubelaker (1994) and subsequently utilized by Walker (2008). Likewise, survey participants indicated that the three traits originally described by Phenice (1969) and then modified by Klales et al. (2012) are the most popular for morphological pelvic sex estimation (Klales 2013). Only a few studies have validated or attempted to recalibrate the predictive sex estimation equations presented in Walker (2008) and Klales et al. (2012) and the results have widely varied (e.g., Roth et al. 2013; Stull et al. 2013; Garvin et al. 2014). Despite being popular and frequently applied to active forensic cases, secular change and population variation in the traits utilized in Walker (2008) and Klales et al. (2012) have not yet been fully explored. Furthermore, many of these morphological traits are bilateral and the effects, if any, of asymmetry have not been evaluated for their impact on sex classification accuracy.
Given the survey results mentioned above, it would appear that most forensic scientists are, in fact, using both types of methods, albeit perhaps with a preference for qualitative methods, for forensic case work. Quantitative sex estimation methods were consolidated into the computer program FORDISC in 1993 (Jantz and Ousley 2005). Cranial and postcranial metric data from 13 modern populations (8 male groups and 5 female groups) have been integrated into the interactive program (Jantz and Ousley 2005). Forensic scientists can enter the measurements of their unknown individual into the program and then compare their case to those individuals within the known reference samples for sex and ancestry classification using linear discriminant function analysis. Users are then provided with discriminant function classification accuracy, a posterior probability of group membership, and several typicalities which the practitioner can then interpret for combined sex and ancestry estimation. Within two short decades, FORDISC has become the number one method for metric assessment of sex and ancestry (Klales 2013). The ease of use and inclusive nature of the program is likely why FORDISC is the number one method for metric assessment of the sex and ancestry parameters of the biological profile (Klales 2013). Unfortunately, a similar program for morphological methods does not yet exist; therefore, forensic scientists must rely on the equations provided in the original Walker (2008) and Klales et al. (2012) publications and then must evaluate sample and statistical appropriateness when applying the method to their unknown individual....until now!
This research was funded by National Institute of Justice Grant 2015-DN-BX-K014. Opinions or points of view expressed in this research represent a consensus of the authors and do not necessarily represent the official position or policies of the U.S. Department of Justice or the National Institute of Justice. Any products and manufacturers discussed are presented for informational purposes only and do not constitute product approval or endorsement by the U.S. Department of Justice or the National Institute of Justice.
