ABSTRACT
The subject of quadratic forms in normal random variables has received a great deal of attention since the early development of the theory of linear models. This is mainly due to the important role quadratic forms play in the statistical analysis of linear models. The sums of squares in a given ANOVA (analysis of variance) table, for example, can each be represented as a quadratic form in the corresponding data vector. These sums of squares are used to develop appropriate test statistics.
