ABSTRACT
Introduction Regression analysis is another way of describing and evaluating relationships between variables. However, unlike correlation there is an assumption that one variable is a variable to be predicted (a DV) and one or more variables (IVs) are used to predict the outcome of the DV. Strictly speaking, the terms DV and IV are more appropriate in experimental research: their equivalents in non-experimental research are criterion variable and predictor variable respectively. However, at the risk of annoying those who prefer the latter terms I am going to use DV and IV throughout this chapter. It allows me to use abbreviations without adding new ones in the form of CV and PV which may introduce their own confusion. Although not one of the factors that affected my decision, it is also consistent with the descriptions used by SPSS. As we will see at the end of the chapter, techniques that analyse designs that look for differences between groups, such as ANOVA, and techniques that analyse designs that are looking for relationships among variables, such as regression, are in fact based on the same principles.
