Illustrates the practical application of the projective approach to linear models. Unlike other books on linear models the use of projections and other vector space ideas does not stop after the general theory; vectors are employed throughout the text. In addition to the standard topics discussed in other books, this book contains discussions of the following: testing for lack of fit in the absence of replication, models with singular convariance matrices, generalized split plot models, mixed models, variance component estimation, best linear prediction, best linear unbiased prediction, residuals, leverage, influential observations, variable selection, collinearity, and loglinear models.
Be the first to review this book!
