Hip Anatomy and Ontogeny of Lower Limb Musculature in Three Species of Nonhuman Primates

Table 2

Mean percentage (SD) of functional muscle groups in adult primates with results of statistical tests.

	Galago	Cheirogaleus	Microcebus	Kruskal-Wallis test
% hip extensors (including hamstrings)	16.0^† (4.1)	27.0 (4.3)	30.0^† (1.7)	𝜒^2 = 8.86, 𝑃<.02
% knee extensors	50.0^† (7.0)	20.0 (2.9)	25.0^† (2.6)	𝜒^2 = 10.38, 𝑃<.01
% ankle plantarflexors	6.0^† (0.6)	7.0 (1.0)	7.0^† (0.6)	𝜒^2 = 7.4, 𝑃<.05

^†Pairs of means that were significantly different using a Mann Whitney U-test with a sequential Bonferroni correction to assess significance [24]. Use of this statistical correction was extensively discussed by Cabin and Mitchell [28]. They noted that failure to use this correction inflates Type I errors (falsely rejecting the null hypothesis), while “overzealous use” of this correction inflates Type II errors (falsely accepting the null hypothesis. One possible approach would be to pool all post hoc tests for correction, in which case the range of corrected 𝑃 values is.006 to.05. In this case, none of the pairs are significantly different, but the likelihood of Type II errors appears markedly increased. We applied the sequential Bonferroni correction separately for the three Mann Whitney U-tests that followed each Kruskal-Wallis test.