Which variance is not utilized in a repeated measures ANOVA?

In the context of a repeated measures ANOVA, the analysis focuses on how the means of different conditions relate to one another while accounting for the repeated observations taken from the same subjects. The analysis measures the variance that can be attributed to the observed differences in conditions as well as the variance attributable to the inherent differences among subjects.

The variance within, which typically refers to the variance among individual scores in the same condition, is not utilized in the same way in a repeated measures ANOVA. Instead, the key components of the repeated measures ANOVA include variance between treatments (how much scores differ from one condition to another), variance residual (which captures the variability not explained by the model), and the total variance (the overall variability in the data).

When analyzing the data in this context, repeated measures allow for the same subjects to be measured across conditions, thus reducing individual differences that can inflate within-group variance. Therefore, while variance within generally reflects error variance in a traditional ANOVA, in a repeated measures design, it is the differences between conditions (variance between) and the error term (variance residual) that are more relevant to the analysis. This distinction makes it clear why variance within is not utilized in the standard framework of a repeated measures ANOVA.