Linear models are central to the practice of statistics. They are part of the core knowledgeexpected of any applied statistician. Linear models are the foundation of a broad range ofstatistical methodologies; this book is a survey of techniques that grow from a linearmodel. Our starting point is the regression model with response y and predictors x1, ... xp. The model takes the form:y=a 0+a 1x1+ ... +a pxp+a where a is normally distributed. This book presents three extensions to this framework. The first generalizes the y part; the second, the a part; and the third, the x part of the linearmodel.