A Comparison of Imputation Methods in a Longitudinal Randomized Clinical Trial. Tang, L., Song, J., Belin, T. R., & Unützer, J. Stat Med, 2005:2111-2128, 2005.
bibtex   
@article{tan05com,
  title = {A Comparison of Imputation Methods in a Longitudinal Randomized Clinical Trial},
  volume = {2005},
  journal = {Stat Med},
  author = {Tang, Lingqi and Song, Juwon and Belin, Thomas R. and Un{\"u}tzer, J{\"u}rgen},
  year = {2005},
  keywords = {missing-data,multiple-imputation,predictive-mean-matching,locf,hot-deck,model-based-imputation},
  pages = {2111-2128},
  citeulike-article-id = {13265438},
  posted-at = {2014-07-14 14:09:56},
  priority = {0},
  annote = {LOCF and available-case method had poor coverage;imputation under a multivariate normal model did not produce correct coverage with highly skewed distributions;hot deck consistently have good nominal coverage and had CL widths 7 per cent larger on average than using multivariate normal imputation;approximate Bayesian bootstrap;different imputation methods used for item and for unit nonresponse;simulation setup;nice graphics;errata 25:1095;2006}
}

Downloads: 0