NORMUL: A FORTRAN program for testing multivariate normality |
| |
Authors: | Ronald R Holden Michael Parent |
| |
Institution: | 1. Department of Psychology, Queen’s University, K7L 3N6, Kingston, ON, Canada
|
| |
Abstract: | NORMUL is a FORTRAN program that provides a test of whether data conform to a multivariate normal distribution. The method involves correlating Mahalanobis distances for observed data with expected chi-square percentile values. This obtained correlation is then tested for significance by empirically evaluating the probability of its belonging to a distribution generated from multivariate normal data. |
| |
Keywords: | |
本文献已被 SpringerLink 等数据库收录! |
|