Missing data are a nearly universal problem for applied researchers, particularly within educational settings where missed responses or losses of participants from a study are common occurrences. Although there are numerous and diverse strategies to address missing data, researchers often use a small collection of strategies (e.g., regression imputation, listwise deletion) that are outdated and inappropriate. The purpose of the current project is to provide researchers with clear research-based solutions to missing data problems that leverage recent methodological advances.

The research team will develop two products that will serve as training tools for educational researchers. First, they will prepare a review paper that provides foundational knowledge about classic missing data topics as well as an up-to-date account of recent methodological innovations; all with a specific focus on application. The paper will provide researchers with a broad view of the methodological literature, accessible rather than mathematical descriptions of missing data handling procedures, and research-based prescriptions for applied practice. They will also produce an instructional video series that illustrates different procedural aspects of a missing data analysis. The video series will provide researchers with step-by-step tutorials that walk through different aspects of missing data analyses using Blimp, a standalone statistical analysis software package that offers sophisticated missing data handling routines—developed under awards R305D150056 and R305D190002. Collectively, the two products will provide education researchers with the resources they need to understand and skillfully utilize cutting-edge missing data methodologies in their own research.

Education researchers will have access to both the review paper and video series through the Applied Missing Data website. The team will also disseminate the video series through YouTube to encourage widespread use by researchers and students worldwide. In addition, the research team will submit the paper for publication and hold workshops to disseminate the work.

Project website: www.appliedmissingdata.com

Related IES Projects: Multiple Imputation Procedures for Multilevel Data (R305D150056); Model-based Multiple Imputation for Multilevel Data: Methodological Extensions and Software Enhancements (R305D190002)