Premium
The statistical analysis of somatotype data
Author(s) -
Cressie N. A. C.,
Withers R. T.,
Craig N. P.
Publication year - 1986
Publication title -
american journal of physical anthropology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.146
H-Index - 119
eISSN - 1096-8644
pISSN - 0002-9483
DOI - 10.1002/ajpa.1330290509
Subject(s) - statistics , multivariate analysis of variance , mathematics , bonferroni correction , pairwise comparison , discriminant function analysis , univariate , null hypothesis , linear discriminant analysis , analysis of variance , statistic , multivariate statistics
The literature contains examples of one‐way ANOVAs being conducted on SAD (somatotype altitudinal distance) values to determine whether there are statistically significant differences between group somatotype means. The problems with this strategy concern the premature collapsing of the three component somatotype vectors into a scalar SAD value together with use of the inappropriate degrees of freedom for the F‐ratio. A theoretical rationale is presented for remedying the latter defect. However, a one‐way MANOVA, which uses Wilks's lambda as the test statistic, is the most powerful method of determining whether there are statistically significant differences between the somatotype means for two or more groups. If the null hypothesis is rejected, then for more than two groups, pairwise comparisons should be conducted by using Hotelling's T 2 with a Bonferroni adjusted alpha level. The one‐by‐one and joint contributions of the somatotype components to each significant pairwise difference can furthermore be determined by univariate F‐ratios, discriminant function analyses, and forward stepwise discriminant analyses.