An Investigation of Factored Regression Missing Data Methods for Multilevel Models with Cross-Level Interactions

A growing body of literature has focused on missing data methods that factorize the joint distribution into a part representing the analysis model of interest and a part representing the distributions of the incomplete predictors. Relatively little is known about the utility of this method for multilevel models with interactive effects. This study presents a series of Monte Carlo computer simulations that investigates Bayesian and multiple imputation strategies based on factored regressions. When the model's distributional assumptions are satisfied, these methods generally produce nearly unbiased estimates and good coverage, with few exceptions. Severe misspecifications that arise from substantially non-normal distributions can introduce biased estimates and poor coverage. Follow-up simulations suggest that a Yeo-Johnson transformation can mitigate these biases. A real data example illustrates the methodology, and the paper suggests several avenues for future research.

Automated summary

A growing body of literature has focused on missing data methods that factorize the joint distribution into a part representing the analysis model of interest and a part representing the distributions of the incomplete predictors. This study explores the utility of these methods for multilevel models with interactive effects, using Monte Carlo computer simulations. The results suggest that these methods generally produce unbiased estimates and good coverage when the model's distributional assumptions are satisfied, but severe misspecifications from non-normal distributions can introduce biases.