Testing Inter ‐ and Intra‐Observer Agreement of the Original and Revised Coimbra Methods

AbstractIn 2009 the Coimbra Method was created to standardized observations of entheseal changes, which are often studied as markers of vigorous physical activity formed during life. Low percentage agreement rates of inter ‐observer replicability led to a revised version of the Coimbra Method published in 2016. Robust statistical significance should be expected of a standardized method, yet neither published method statistically tested inter‐ and intra‐observer agreement. This paper tests reliability of both Coi mbra Method versions using Cohen’s Kappa, and percentage of agreement for comparison to the originally published rates. This paper also marks the first test of reliability for the Coimbra Method that is independent of the original authors. Fibrocartilaginous entheses of the humerus, radius, and sc apula were observed on 82 individuals from Scottish and Cretan collections. Results ranged from high intra‐observer agreement (86.9 to 91.8%) to lower inter‐observer agreement (70.5 to 76.7%) for both methods. Inter‐observer agreement only slightly improved between the original (71.8%) and rev ised (75.2%) Coimbra Methods. Results for Kappa ranged from substantial for intra‐observer agreement (k=0.70), but only moderate for inter‐observer agreement (k=0.57) for both methods. Either 80% or ‘substantial’ kappa agreement is considered the minimum for a standardized method, and these results indicate that inter‐observer agreement from even the revised Coi...
Source: International Journal of Osteoarchaeology - Category: Science Authors: Tags: RESEARCH ARTICLE Source Type: research