Bootstrap tests of significance and the case for humanlike skeletal-size dimorphism in Australopithecus afarensis

Most estimates of sexual size dimorphism in Australopithecus afarensis indicate that this early hominin was more dimorphic than modem humans. In contrast, a recent study reported that size variation in A. afarensis, as represented by postcranial remains from Hadar and Maka, Ethiopia, is statistically most similar to that of modem humans, indicating a humanlike level of sexual dimorphism. Here, we evaluate the evidence for humanlike dimorphism in A. afarensis. We argue that statistical support for this claim is not as robust as has been asserted for the following reasons: (1) the analysis from which the claim was derived does not distinguish the A. afarensis sample from either the human or chimpanzee samples; (2) for some of the comparisons made, the A. afarensis sample cannot be distinguished from the Gorilla sample using two-tailed tests; and (3) the A. afarensis postcranial sample used in the analysis may contain more male than female specimens, which precludes a straightforward interpretation of the statistical results. Thus, support for humanlike dimorphism is equivocal, and a greater level of dimorphism cannot be ruled out. (c) 2006 Elsevier Ltd. All rights reserved.