Fig. 2From: Evaluation of approaches for multiple imputation of three-level dataEmpirical standard errors (filled circles with error bars showing ±1.96× Monte Carlo standard errors) and average model-based standard errors (hollow circles) from 1000 simulated datasets, for available case analysis (ACA) and the 8 multiple imputation (MI) approaches under two scenarios for missing data proportions at waves 2,4 and 6 (10%, 15%, 20% and 20%, 30%, 40%, respectively) and four ICC combinations when data are missing at random (MAR-CATS). The following abbreviations are used to denote different MI methods, e.g., DI: dummy indicators, FCS: fully conditional specification, JM: joint modellingBack to article page