Contents Index <<Browse<< >>Browse>>
Here proc glm is used to perform multivariate analysis of variance (MANOVA).
The Fisher Iris Data, introduced in Project 3 and contained in the file iris.dat, is analyzed.
1. Use the manova statement in proc glm, as described in Proc GLM for MANOVA to test the hypothesis that the vector of 4 measurements of each species of iris have the same mean vector. What is your conclusion? Turn in print outs of only the relevant parts of your analysis, and also of your program. What are the implications of your analysis on the prospect of using discriminant analysis based on the 4 measurements to distinguish between species?
2. Test the hypothesis that the length measurements are twice the width measurements for both sepal and petal. Is this hypothesis accepted? Test the hypothesis that the sepal length means are the same for all 3 species. Is this the same as the univariate one way ANOVA of Project 8? Turn in suitable printouts to support your answers. Also turn in a copy of your program.
3. An experimenter believes that the sepal width is a linear function of the other 3 measurments. Test this hypothesis; describe your methods and conclusions. Turn in suitable printouts to support your answers and also a copy of your program.
Contents Index <<Browse<< >>Browse>>
Copyright © 1997 by Jerry Alan Veeh. All rights reserved.