11 November 2018 0 6K Report

Generally, there are three assumptions for ANOVA model

1. Independence of cases

2. Normality

3. Equality (or "homogeneity") of variances

What will happen if the cases in each group are dependent? Will the p value for each parameter in this model be overestimated or underestimated? Or both is possible?

Similar questions and discussions