Abstract
Femoral head breadth is widely used in body mass estimation in biological anthropology. Earlier research has demonstrated that reduced major axis (RMA) equations perform better than least squares (LS) equations.
Although a simple RMA equation to estimate body size from femoral head breadth is sufficient in most cases, our experiments with male skeletons from European data (including late Pleistocene and Holocene skeletal samples) and the Forensic Anthropology Data Bank data (including the W.M. Bass Donated Skeletal Collection sample) show that including femoral length or anatomically estimated stature in an equation with femoral head breadth improves body mass estimation precision.
More specifically, although directional bias related to body mass is not reduced within specific samples, the total estimation error range, directional bias related to stature, and temporal fluctuation in estimation error are markedly reduced. The overall body mass estimation precision of individuals representing different temporal periods and ancestry groups (e.g., African and European ancestry) is thus improved. (C) 2017 Published by Elsevier Ltd.