Vertaile menetelmiä

	Bootstrap-simulaatio puuttuvalla datalla	Gibbs-otanta puuttuvalla datalla
Tieteenala	Bayesilainen tilastotiede	Bayesilainen tilastotiede
Menetelmäperhe	Bayesian methods	Bayesian methods
Syntyvuosi	1979–1990s	1987–1990
Kehittäjä	Bradley Efron (bootstrap); missing-data extensions by Efron, Little, Rubin and others	Tanner & Wong (data augmentation), Gelfand & Smith (Gibbs sampler)
Tyyppi	Resampling simulation	Bayesian computational method
Alkuperäislähde	Efron, B. & Tibshirani, R. J. (1993). An Introduction to the Bootstrap. Chapman and Hall/CRC. ISBN: 978-0412042317	Tanner, M. A. & Wong, W. H. (1987). The calculation of posterior distributions by data augmentation. Journal of the American Statistical Association, 82(398), 528–540.
Rinnakkaisnimet	bootstrap with missing data, bootstrap imputation simulation, resampling under missingness, bootstrap MI	data augmentation Gibbs sampler, Gibbs sampler with data augmentation, Bayesian imputation via Gibbs sampling, MCMC missing data imputation
Liittyvät	5	6
Tiivistelmä	Bootstrap simulation with missing data combines resampling-based variance estimation with principled handling of incomplete observations. Rather than deleting cases or assuming complete data, the method integrates imputation or weighting directly into the bootstrap loop, propagating the additional uncertainty due to missingness into the final standard errors and confidence intervals.	Gibbs sampling with missing data treats unobserved values as additional unknowns alongside model parameters and samples all of them jointly within a Markov chain Monte Carlo loop. The method alternates between drawing the missing values from their conditional distribution given the parameters and drawing the parameters from their conditional distribution given the completed data, producing a posterior over both simultaneously.