Handling Missing Data in Cross-Classified Multilevel Analyses: An Evaluation of Different Multiple Imputation Approaches

Abstract

Multiple imputation (MI) is a popular method for handling missing data. In education research, it can be challenging to use MI because the data often have a clustered structure that need to be accommodated during MI. Although much research has considered applications of MI in hierarchical data, little is known about its use in cross-classified data, in which observations are clustered in multiple higher-level units simultaneously (e.g., schools and neighborhoods, transitions from primary to secondary schools). In this article, we consider several approaches to MI for cross-classified data (CC-MI), including a novel fully conditional specification approach, a joint modeling approach, and other approaches that are based on single- and two-level MI. In this context, we clarify the conditions that CC-MI methods need to fulfill to provide a suitable treatment of missing data, and we compare the approaches both from a theoretical perspective and in a simulation study. Finally, we illustrate the use of CC-MI in real data and discuss the implications of our findings for research practice.

An Individual Participant Meta-Analysis to Support Power Analyses for Randomized Intervention Studies in Preschool: Cognitive and Socio-Emotional Learning Outcomes

“…The percentage of missing values per variable varied from 0% to 45.3% (Table 1). To deal with missing data we used an adjusted cluster-means imputation approach for multilevel data (Grund et al, 2023) and generated 50 multiply imputed datasets for each sample using the mice (van Buuren & Groothuis-Oudshoorn, 2011) and miceadds ) R packages. Notably, the applied imputation model was compatible with the models that we specified for estimating single-and multilevel design parameters.…”

The Relationship Between Math Anxiety and Math Achievement: Novel and Robust Evidence From A Meta-Analysis of Individual Participant Data From International Large-Scale Assessmentsvidual Participant Data and Aggregated Data in a Meta-Analysis

“…Handling of Missing Data. In the first step of our effect-size calculations, we applied multiply nested imputations to impute the missing data in our target variables (Weirich et al, 2014) while taking into account the nested structure of the data (Grund et al, 2023). The details of the imputation process can be found in OSM 4.…”