ABSTRACT
In chapter 4 you learned how to analyze a two-factor completely randomized ANOVA. In this chapter, those commands are extended to a three-factor design (and beyond).
The data for this example are from Stevens (1999, p. 174). In order to make the output more understandable, the variables have been renamed from Stevens’ original terms. In this study, the researcher investigated the effects of grade and gender of the participant and type of feedback on memory. The dependent variable is ‘score’, the participant’s score on a memory test. There are three independent variables: Factor A is grade (either first grade, coded 1, or third, coded 2), Factor B is the gender of the participant (either boy, coded 1, or girl, coded 2), and Factor C is the type of feedback a person received (either positive, coded 1, negative, coded 2, or none, coded 3). This design is a 2 × 2 × 3 analysis of variance. Each participant of a certain level of age (‘facta’) and a certain gender (‘factb’) received a certain level of feedback (‘factc’). For example, one group of participants was male first graders receiving positive feedback.There are 12 (= 2 × 2 × 3) such groups of participants in all.
