Print Email Facebook Twitter Multiple Imputation of Predictor Variables Using Generalized Additive Models Title Multiple Imputation of Predictor Variables Using Generalized Additive Models Author de Jong, R. van Buuren, S. Spiess, M. Publication year 2016 Abstract The sensitivity of multiple imputation methods to deviations from their distributional assumptions is investigated using simulations, where the parameters of scientific interest are the coefficients of a linear regression model, and values in predictor variables are missing at random. The performance of a newly proposed imputation method based on generalized additive models for location, scale, and shape (GAMLSS) is investigated. Although imputation methods based on predictive mean matching are virtually unbiased, they suffer from mild to moderate under-coverage, even in the experiment where all variables are jointly normal distributed. The GAMLSS method features better coverage than currently available methods. Subject LifeCH - Child HealthELSS - Earth, Life and Social SciencesHealthy for LifeHealthy LivingComparison of imputation methodsGeneralized additive models for locationLinear modelMultiple imputationPredictive mean matchingRobust imputation modelsScale and shape To reference this document use: http://resolver.tudelft.nl/uuid:c654384f-5cc9-45bf-8bb6-349348f23b92 DOI https://doi.org/10.1080/03610918.2014.911894 TNO identifier 844048 Source Communications in Statistics Simulation and Computation, 45 (3), 968-985 Document type article Files To receive the publication files, please send an e-mail request to TNO Library.