Different approaches to defining suppression in multiple regression/correlation are compared, and their differences are illustrated. A test for determining the significance of a suppressor effect, which is based on the definition of suppression of W. F. Velicer, is extended to the general multiple predictor case and analysis of variance. (SLD)

The suppressor variable, a variable wholly uncorrelated with a criterion, but which nevertheless improves prediction because of its relationship with a predictor, is critically examined. For a suppressor so defined, formal identities are shown with part, partial, and multiple correlational procedures. It is demonstrated that if maximum prediction…

Describes a study of 38 students at Eastern Oregon State College which identified mathematics proficiency as the key student characteristic leading to successful completion of computer science courses, and developed a model that accounts for 82 percent of the variation in students' final grades in computer courses for use by advisers. (LRW)