What does "Residual" variance refer to in ANOVA?

"Residual" variance in ANOVA represents the portion of the total variance that is not explained by the treatment effects or the model. It quantifies the variability in the data that remains after accounting for the differences between the treatment groups. In other words, it reflects how much individual data points deviate from their group means—indicating random error or inherent variability among observations.

Understanding residual variance is crucial because it helps researchers assess the effectiveness of their model. If the residual variance is large compared to the variance explained by the model, it suggests that the treatment effects may not be significant.

The other options pertain to different aspects of variance in ANOVA, such as variance explained by treatments or the variability within groups, which are not representative of the residual variance itself. Thus, "Residual" variance is specifically tied to the unexplained variability in the context of the model.