Estimation of stature in archaeological human skeletal remains from Britain.

OBJECTIVES: For archaeological populations of northwest European origin, the Trotter and Gleser () "White" equations are generally used to estimate stature from long-bone length. The aim of this research is to investigate the reliability of these formulae for this purpose.

MATERIALS AND METHODS: Stature estimated using the Trotter and Gleser formulae was evaluated by comparing estimates with anatomically reconstructed stature in adult skeletons (N = 40) from Mediaeval England. Performance of the Trotter and Gleser equations based on femur length was compared (using standard error of the estimate (SEE) and mean percentage error (%error)) with others that have been derived from archaeological and recent European ancestry populations. Performance of ordinary least squares (OLS) and reduced major axis (RMA) versions of some of the equations was also evaluated.

RESULTS: 14 male and 11 female femoral equations were investigated. For females, the original (OLS) Trotter and Gleser "White" equations produced the lowest SEE and %error. For the males, the most reliable was the RMA version of the Trotter and Gleser "White" WWII equation; the Trotter and Gleser equations based on the Korean War dead performed poorly.

DISCUSSION: In an archaeological target population, stature estimation equations devised using archaeological reference populations from a similar ecogeographic zone did not perform better than those based on recent populations. There was no indication that either the RMA or OLS approach produced consistently lower SEE and %error, but the former more faithfully represented the variation in anatomically reconstructed stature present in the study group.