This paper focuses on the Multivariate Analysis of Variance (MANOVA), a method widely used in the biosciences. Engineers involved in experimental work can find linear discriminants, ellipsoidal contours, and Hate/ling's P test useful when they compare two groups of experiments that differ by some criteria. With ellipsoidal contours, they can visualize the approximate extent of the distribution of the data in each group. With Hate/ling 's T2 test, they can determine whether the differences between the groups are occurring by more than chance. With linear discriminant functions, they can define a model to determine the membership of a data point to a group. Such a model is referred to as a classifier. Two numerical examples are presented with the use of graphical illustrations. The first example explains the presentation of the three methods. The second example utilizes experimental data on pull-out tests of prestressing strands.
