ABSTRACT

Designing experiments allows one to properly characterize variation of the conditions under consideration. Experimental design is a broad area in Statistics and provides researchers with tools, strategies, and knowledge to avoid shortcomings that could arise in the analysis stage. Most models in experimental design are also referred to as ANOVA models because the objective is to characterize significant sources of variation. The majority of these models are linear models and can also be written using a linear regression model. Experimental design is a very extensive topic, so this chapter discusses only some specific models where regression plays a more prominent role. Some classical applied texts on experimental design include Kuehl (2000), Box et al. (2005), and Montgomery (2013). A couple of theoretical texts on the topic are Pukelsheim (1993) and Bailey (2008).