ANOVA / ANCOVA

• Balanced and unbalanced designs
• Missing cells
• Factorial, nested, Latin-square, and mixed designs
• Repeated measures
• Box, Greenhouse–Geisser, and Huynh–Feldt corrections

Postestimation after ANOVA

• Tables of adjusted means
  • Tests for effects, including pooling and nonresidual error terms
  • Tests for expressions involving the coefficients of the underlying regression model
• Bonferroni, Holm, and Šidák adjustments for multiple tests
• Ability to display symbolic forms
• Predictions and influence statistics
  • Expected values
  • Residuals, standardized residuals, studentized residuals
  • Standard error of the prediction or residuals
  • Leverage
  • Cook’s D
  • COVRATIO
  • DFBETAs
  • Diagonal of hat matrix
  • Welsch distance
• Linear combinations of estimated coefficients along with confidence intervals
• Diagnostic plots
  • Component plus residual
  • Augmented component plus residual
  • Leverage vs. squared residual
  • Residual vs. fitted
  • Residual vs. predictor

MANOVA

• Multivariate test statistics
  • Wilks’s lambda
  • Pillai’s trace
  • Lawley–Hotelling trace
  • Roy’s largest root
• Balanced and unbalanced designs
• Missing cells
• Factorial, nested, Latin-square, and mixed designs
• Repeated measures

Postestimation after MANOVA

• Tables of adjusted means
• Multivariate tests (Wilks’s lambda, Pillai’s trace, etc.) for
  • Terms from the model
  • Pooled terms
  • Terms (or pooled terms) tested using other terms (or pooled terms) as the error term
  • Linear combinations of the underlying design matrix
• Wald tests of expressions involving the coefficients of the underlying regression model
• Linear combinations of estimated coefficients along with confidence intervals
• Predictions
  • Point estimates
  • Standard error of point estimates
  • Residuals