Computation of Effect Size for Moderating Effects of Categorical Variables in Multiple Regression
Abstract
Computer Program Exchange Computation of Effect Size for Moderating Effects of Categorical Variables in Multiple Regression Herman Aguinis, Business School, University of Colorado at Denver and Health Sciences Center Charles A. Pierce, Fogelman College of Business & Economics, University of Memphis The computation and reporting of effect size estimates is becoming the norm in many journals in psychology and related disciplines (Kendall, 1997; Thompson, 1994; Zedeck, 2002). Despite the increased importance of effect sizes, researchers may not report them or may report inaccurate values because of a lack of appropriate computational tools. For instance, Pierce, Block, and Aguinis (2004) provided examples of articles published in prestigious journals such as Psychological Science, Developmental Psychology, Journal of Educational Psychology, and Journal of Abnormal Psychology, in which researchers erroneously reported partial eta-squared values as representing classical eta-squared values. One likely reason for why researchers do not report effect size estimates or, even worse, report inaccurate values is that most commercially available statistical software packages provide only a limited number of effect-size estimates (e.g., 2 but not f2 Þ. The use of multiple regression to estimate moderating (i.e., interaction) effects of categorical variables involves an ordinary least squares (OLS) regression equation that