Inter-Observer Variation in Methodologies Involving the Pubic Symphysis, Sternal Ribs, and Teeth

• Published in: Journal of forensic sciences Vol. 53; no. 3; pp. 594 - 600
• Main Authors: Kimmerle, Erin H., Prince, Debra A., Berg, Gregory E.
• Format: Journal Article
• Language: English
• Published: Oxford, UK Blackwell Publishing Ltd 01.05.2008; Wiley Subscription Services, Inc
• Subjects: Age Determination by Skeleton - methods; Age Determination by Teeth - methods; aging methods; Balkans; Bones; Dentin - pathology; Female; Forensic Anthropology - methods; Forensic Dentistry - methods; Forensic odontology; Forensic osteology; forensic science; Human remains; Humans; identification; Male; Methods; Observer Variation; Pubic Symphysis - anatomy & histology; repeatability; Reproducibility of Results; Ribs - anatomy & histology; Teeth; Tooth - anatomy & histology; Tooth Root - anatomy & histology; Kosovo
• ISSN: 0022-1198; 1556-4029
• DOI: 10.1111/j.1556-4029.2008.00715.x

:  For the skeletal age of a victim to be useful in victim identification, the methods on which it is based must be reliable, accurate, and the results easily duplicated. The ability of multiple investigators to duplicate results is an interesting and complex issue. The purpose of this study is to investigate how consistently multiple investigators assign skeletal traits to rib, pubic symphyseal, or tooth “phases” and measure teeth. The skeletal data from identified individuals in Kosovo are used to test inter‐observer variation for a variety of skeletal and dental aging techniques. Two hundred and ninety‐six (n = 296) pubic symphyses were scored in the manners of the Todd’s ten‐phase system and the Suchey‐Brooks six‐phase system. Six hundred and twenty‐two (n = 622) sternal rib ends were scored in the manner of İşcan and co‐author’s nine‐phase system. Four hundred and twelve (n = 412) single‐rooted teeth were measured in the manner of Lamendin and colleagues and scored for the amount of tooth wear using Smith’s nine‐phase system. Repeat measures were taken by multiple observers. There appears to be a wide range of variation, even among experienced investigators in the assignment of phase or metric data. Inter‐observer variation, investigated through Pearson’s r correlation coefficients, the Wilcoxon signed ranks test, and paired samples t‐tests demonstrate significant differences using all methods. How this variation affects the accuracy of age estimation is subject to further investigation, but what is clear is that even with collaboration among investigators to calibrate with one another, the repeatability of numerous aging methodologies is difficult to achieve. Through this investigation it appears the problem lies in the qualitative nature of broad descriptive phase categories, which contain multiple skeletal features and traits that are open to interpretation.