Machine-Learning Models for Handling Data-Missingness in Educational Research: A Comprehensive Review

• In Event: AERA Roundtable Session Thursday 1:45 pm Four Seasons Ballroom 4
  In Roundtable Session: Machine Learning Techniques II: Advancing Coding/Essay Scoring and Estimation of Treatment Effects (Table 2)

Abstract

Researchers are usually faced with the challenge of dealing with incomplete datasets. The missing data imputation methods can affect the validity of data analysis results. Machine learning (ML) models for multiple imputation are recommended in the literature to improve the quality of data imputation. Concerning the unfamiliarity of many educational researchers on ML models for missing data imputation, this study conducted a comprehensive review on the use of ML algorithms for educational studies when dealing with data-missingness by examining literature published between 2019 and 2023. A comprehensive and synthetic summary of ML models used in educational research is provided. This study will provide educational researchers with an understanding of the existing ML imputation methods and recommendations for best practices.

Authors

• Comfort Happiness Omonkhodion, University of Central Florida

• Haiyan Bai, University of Central Florida

• Oluwaseun Peter Farotimi, University of Tennessee