Re: Gender by Tarsal!
We were finding differentiation by sex in human tarsal elelements 20 years ago in the Oxnard Lab of Western Australia. Their comment that "[roughly]... one dimension can be misleading, however if one uses two..." is a simplistic way of saying that there is a clear case for a multivariate analysis. We found that, using simple principal components analysis, on appropriately transformed data, a clear sexual dimorphism in the talus, calcaneus and cuboid. The latter was particularly the case in certain groups from Southern Africa. Multivariate analysis is an incredibly powerful tool, when used appropriately. In fact sex was a minor player; far greater patterns of morphological variation found using this and the similar but more powerful canonical variates analysis, were related to shape. Clearly, most of this was related to function - but there was frequently a small but finite amount related to phylogeny, "gene pool", geography, or similar quantum.
|