Journal of the Korean Data and Information Science Society

The Korean Data and Information Science Society (한국데이터정보과학회)
권호 표지

Comparison of EM with Jackknife Standard Errors and Multiple Imputation Standard Errors

  • Kang, Shin-Soo (Department of Information Statistics, Kwandong University)
  • Published : 2005.11.30
Download PDF
⟨ Previous Next ⟩

Abstract

Most discussions of single imputation methods and the EM algorithm concern point estimation of population quantities with missing values. A second concern is how to get standard errors of the point estimates obtained from the filled-in data by single imputation methods and EM algorithm. Now we focus on how to estimate standard errors with incorporating the additional uncertainty due to nonresponse. There are some approaches to account for the additional uncertainty. The general two possible approaches are considered. One is the jackknife method of resampling methods. The other is multiple imputation(MI). These two approaches are reviewed and compared through simulation studies.

Keywords

References

  1. Journal of the Royal Statistical Society Series B v.39 Maximum likelihood from incomplete data via the EM algorithm(with discussion) Dempster, A.P.;Laird, N.M.;Rubin, D.B.
  2. Statistical Analysis With Missing Data Little, R.J.A.;Rubin, D.B.
  3. Biometrika v.63 Inference and missing data(with discussion) Rubin, D.B.
  4. Proceedings of the Survey Research Methods Section Multiple imputation in sample surveys - A phenomenological bayesian approach to nonresponse Rubin, D.B.
  5. Journal of Korean Data & Information Science Society v.16 no.1 Comparison of EM and multiple imputation methods with traditional methods in monotone missing pattern Kang, S.S.
  6. S-Plus 6.1 Manual, Analyzing Data With Missing Values in S-Plus Insightful Corporation