ABSTRACT
This chapter presents a selection of regression procedures which are useful in experimental design problems. It deals with a topic perhaps unfamiliar to some readers called orthogonal polynomials. The chapter examines various measures concerned with the utility of the assumed regression model in making estimates and predictions. It uses information contained in statistical analysis system printouts in order to demonstrate basic inference procedures. The chapter introduces the notation that provides us with a "shorthand" way of revealing the procedural origin of a sum of squares quantity. This notation will be especially helpful later in this text as a concise means of presenting ANOVA sums of squares "formulas" in the analysis of experimental design data. In the analysis of experimental design data, the investigator is seldom interested in the coefficients of the fitted function.
