ABSTRACT
This chapter continues the three-chapter sequence on analysis of variance (ANOVA) with a discussion of repeated-measures ANOVA. The chapter begins by explaining that repeated-measures ANOVA involves comparing the means of multiple variables within a single sample. Analogies to the dependent-samples t test are discussed, as are advantages of repeated-measures ANOVA over dependent t tests in certain situations. Next, through an extended example, a consideration of when to use different kinds of repeated-measures ANOVA (e.g., alone, with a covariate, and with one or more between-subjects factors) is offered, along with the concepts of within-subjects and between-subjects effects. The in-depth section of this chapter includes discussions of key concepts, including how variance is partitioned into the within-subjects component (e.g., variances in the means attributable to change over time or trials), the error component, the between-subjects component, and effects of the covariates. A detailed example is provided to illustrate how repeated-measures ANOVA works, including output from an SPSS analysis that includes the effect size, partial eta squared. The chapter continues with an example of a repeated-measures analysis of covariance (ANCOVA), and then an example of a mixed-model ANOVA that includes within-subjects and between-subjects factors. Additional examples are provided to illustrate interactions involving within-subject and between-subject effects and how to interpret them. The chapter concludes with a detailed analysis of SPSS output for a mixed-model ANOVA and a caution about misleading graphs.
