Evaluation of Missing Data Analytical Techniques in Longitudinal Research: Traditional and Machine Learning Approaches

arXiv:2406.13814v1 Announce Type: cross
Abstract: Missing Not at Random (MNAR) and nonnormal data are challenging to handle. Traditional missing data analytical techniques such as full information maximum likelihood estimation (FIML) may fail with nonnormal data as they are built on normal distribution assumptions. Two-Stage Robust Estimation (TSRE) does manage nonnormal data, but both FIML and TSRE are less explored in longitudinal studies under MNAR conditions with nonnormal distributions. Unlike traditional statistical approaches, machine learning approaches do not require distributional assumptions …

abstract arxiv assumptions data distribution evaluation fail information likelihood machine machine learning maximum maximum likelihood estimation normal random research robust stage stat.ap stat.me stat.ml type