ABSTRACT
INTRODUCTION Multiple regression focuses on the prediction of a criterion variable (also known as dependent variable, outcome variable, response variable) from two or more predictors (also known as independent variables, regressors). The criterion must be continuously scaled. Predictors may be continuously scaled or dichotomous. Predictors with three or more categories are converted to a set of dichotomous predictors via dummy coding (see Cohen, Cohen, West, & Aiken, 2003). This chapter presents power analyses for the R2 model, R2 change, and regression coefficients in designs using multiple predictors. In addition, the chapter includes tests that examine differences between independent and dependent predictors as well as tests comparing R2 across independent samples. The “Additional Issues” section includes discussions of the impact of reliability on power and detecting multiple effects.
