Typicality and Predictive Distributions in Discriminant Function Analysis

While discriminant function analysis is an inherently Bayesian method, researchers attempting to estimate ancestry in human skeletal samples often follow with the calculation of frequentist-based typicalities for assigning group membership. Such approach problematic because it fails account admixture and variation why individuals may be classifijied as outliers or nonmembers particular groups. This article presents argument methodology employing a fully applied cases estimation. The requires adding calculation, estimation, predictive distributions fijinal step ancestry-focused analyses. methods multivariate are illustrated using craniometrics from identifijied population within Howells published data. also ways visualize calculated more than three dimensions, explains limitations typicality measures, suggests analytical route future studies based analysis.