EUCLIDEAN DISTANCE MATRIX ANALYSIS - CONFIDENCE-INTERVALS FOR FORM AND GROWTH DIFFERENCES

Analysis of biological forms using landmark data has received substant ial attention recently. Much of the statistical work in this area has concentrated on the estimation of average form, average form differenc e, and average growth difference. From the statistical, as well as the scientific point of view, it is important that any estimate of a scie ntifically relevant quantity be accompanied by a statement regarding i ts accuracy. Such a statement is contained in a confidence interval. T he purpose of this paper is to provide a method to obtain confidence i ntervals for form difference and growth difference estimators. The est imators are based on Euclidean distance matrix analysis. The confidenc e intervals are calculated using the model independent bootstrap metho d. We illustrate the method by using three examples: morphological dif ferences between samples of craniofacial patients and normal controls using two dimensional data from head X-rays, sexual dimorphism of cran iofacial morphology in Cebus apella, and sexual dimorphism of facial g rowth in Cebus apella using three-dimensional data. (C) 1995 Wiley-Lis s, Inc.