Multiple imputation in three or more stages

Abstract

Missing values present challenges in the analysis of data across many areas of research. Handling incomplete data incorrectly can lead to bias, over-confident intervals, and inaccurate inferences. One principled method of handling incomplete data is multiple imputation. This article considers incomplete data in which values are missing for three or more qualitatively different reasons and applies a modified multiple imputation framework in the analysis of that data. Included are a proof of the methodology used for three-stage multiple imputation with its limiting distribution, an extension to more than three types of missing values, an extension to the ignorability assumption with proof, and simulations demonstrating that the estimator is unbiased and efficient under the ignorability assumption.

An examination of ethnic identity as a mediator of the effects of community participation and neighborhood sense of community on psychological empowerment among urban youth of color

“…The largest amount of missing data in this study was associated with ethnic identity (10%). Although various missing data techniques are available (McGinniss & Harel, ), missing data were handled using maximum likelihood (ML) estimations through AMOS (version 23.0) software, which addresses the missing data and parameter estimates, and estimates the standard error in a single step (Hancock & Liu, ). Using AMOS to handle missing data also allows for a theoretically informed direct approach to handling missing data through modeling, opposed to other imputation methods, which can be designated as indirect (Byrne, ).…”

Validation of the Brief Sense of Community Scale among youth of color from an underserved urban community

“…Further inspection of these data revealed that the largest amounts of missing data were related to school importance (< 10%). Although, numerous missing data techniques are available (McGinniss & Harel, ), missing data for this study were handled using maximum likelihood through IBM SPSS AMOS (v. 25), a structural equation modeling software.…”

Validation of an abbreviated Sociopolitical Control Scale for Youth among a sample of underresourced urban youth of color

“…Further inspection of data revealed that no more than 5% of data were missing for any given variable. Although numerous missing data techniques are available (McGinniss & Harel, ), missing data for this study were handled using maximum likelihood (ML) through AMOS SEM software.…”