1. Multiple Imputation to Balance Unbalanced Designs for Two-Way Analysis of Variance
- Author
-
Pieter M. Kroonenberg and Joost R. van Ginkel
- Subjects
multiple imputation ,Pooling ,Two-way analysis of variance ,01 natural sciences ,type-iii sum of squares ,010104 statistics & probability ,missing data ,0504 sociology ,Statistics ,Statistics::Methodology ,Psychology ,0101 mathematics ,Unbalanced data ,General Psychology ,Mathematics ,Statistics::Applications ,05 social sciences ,Explained sum of squares ,050401 social sciences methods ,General Social Sciences ,Missing data ,Power (physics) ,BF1-990 ,two-way analysis of variance ,Analysis of variance ,unbalanced designs - Abstract
A balanced ANOVA design provides an unambiguous interpretation of the F-tests, and has more power than an unbalanced design. In earlier literature, multiple imputation was proposed to create balance in unbalanced designs, as an alternative to Type-III sum of squares. In the current simulation study we studied four pooled statistics for multiple imputation, namely D₀, D₁, D₂, and D₃ in unbalanced data, and compared them with Type-III sum of squares. Statistics D₁ and D₂ generally performed best regarding Type-I error rates, and had power rates closest to that of Type-III sum of squares. Additionally, for the interaction, D₁ produced power rates higher than Type-III sum of squares. For multiply imputed datasets D₁ and D₂ may be the best methods for pooling the results in multiply imputed datasets, and for unbalanced data, D₁ might be a good alternative to Type-III sum of squares regarding the interaction.
- Published
- 2021