SEDL / STP / STP652-EB / STP35658S

A Multivariate Procedure for Comparing Control Areas to Possible Impact Areas, Graphically Displaying the Results

Callahan, JD
Assistant professor of Statistics, University of Rhode Island, Kingston, R. I.

Palmer, J
Research scientist, Southern California Edison Company, Rosemead, Calif

Pages: 14    Published: Jan 1978

Download this paper for $25 PDF (148K)          View License Agreement
Abstract

Principal components are used to reduce the dimension of highly correlated, multivariate data. After applying analyses of variance to the principal components, looking for site differences, a graphical method of multiple comparisons is presented. The procedure is applied to two large sets of benthic data.