×

Estimated conditional score function for missing mechanism model with nonignorable nonresponse. (English) Zbl 06841771

Summary: Missing data mechanism often depends on the values of the responses, which leads to nonignorable nonresponses. In such a situation, inference based on approaches that ignore the missing data mechanism could not be valid. A crucial step is to model the nature of missingness. We specify a parametric model for missingness mechanism, and then propose a conditional score function approach for estimation. This approach imputes the score function by taking the conditional expectation of the score function for the missing data given the available information. Inference procedure is then followed by replacing unknown terms with the related nonparametric estimators based on the observed data. The proposed score function does not suffer from the non-identifiability problem, and the proposed estimator is shown to be consistent and asymptotically normal. We also construct a confidence region for the parameter of interest using empirical likelihood method. Simulation studies demonstrate that the proposed inference procedure performs well in many settings. We apply the proposed method to a data set from research in a growth hormone and exercise intervention study.

MSC:

62G05 Nonparametric estimation
62N01 Censored data models
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] Dempster A P, Laird N M, Rubin D B. Maximum likelihood from incomplete data via the em algorithm. J Roy Statist Soc B, 1977, 39: 1-38 · Zbl 0364.62022
[2] Fan J, Gijbels I. Local Polynomial Modeling and Its Applications. London: Chapman and Hall, 1996 · Zbl 0873.62037
[3] Fan J, Yao Q. Nonlinear Time Series: Nonparametric and Parametric Methods. New York: Springer-Verlag, 2003 · Zbl 1014.62103 · doi:10.1007/b97702
[4] Hall P, Murison R D. Correcting the negativity of high-order kernel density estimators. J Multivariate Anal, 1993, 47: 103-122 · Zbl 0778.62034 · doi:10.1006/jmva.1993.1073
[5] Härdle W, Stoker T M. Investigating smooth multiple regression by method of average derivatives. J Amer Statist Assoc, 1989, 84: 986-995 · Zbl 0703.62052
[6] Kenward M, Molenberghs G, Thijs H. Pattern-mixture models with proper time dependence. Biometrika, 2003, 90: 53-71 · Zbl 1035.62112 · doi:10.1093/biomet/90.1.53
[7] Kiel D, Puhl J, Rosen C, et al. Lack of association between insulin-like growth factor-I and body composition, muscle strength, physical performance or self reported mobility among older persons with functional limitations. J Amer Geriatr Soc, 1998, 46: 822-828 · doi:10.1111/j.1532-5415.1998.tb02714.x
[8] Kim J K, Yu C Y. A semi-parametric estimation of mean functionals with non-ignorable missing data. J Amer Statist Assoc, 2011, 106: 157-165 · Zbl 1396.62032 · doi:10.1198/jasa.2011.tm10104
[9] Little R J A, Rubin D B. Statistical Analysis with Missing Data, 2nd ed. New York: Wiley, 2002 · Zbl 1011.62004
[10] Liu D P, Zhou X H. A model for adjusting for nonignorable verification bias in estimation of the roc curve and its area with likelihood-based approach. Biometrics, 2010, 66: 1119-1128 · Zbl 1233.62183 · doi:10.1111/j.1541-0420.2010.01397.x
[11] Mack Y, Silverman B. Weak and strong uniform consistency of kernel regression estimates. Probab Theory Related Fields, 1982, 61: 405-415 · Zbl 0495.62046
[12] Owen A B. Empirical likelihood ratio confidence regions. Ann Statist, 1990, 18: 90-120 · Zbl 0712.62040 · doi:10.1214/aos/1176347494
[13] Qin J, Leung D, Shao J. Estimation with survey data under nonignorable nonresponse or informative sampling. J Amer Statist Assoc, 2002, 97: 193-200 · Zbl 1073.62513 · doi:10.1198/016214502753479338
[14] Rubin D. Multiple Imputation for Nonresponse in Surveys. New York: John Wiley & Sons, 1987 · Zbl 1070.62007 · doi:10.1002/9780470316696
[15] Rubin D. Multiple imputation after 18+ years (with discussion). J Amer Statist Assoc, 1996, 91: 473-489 · Zbl 0869.62014 · doi:10.1080/01621459.1996.10476908
[16] Schafer J L. Analysis of Incomplete Multivariate Data. London: Chapman and Hall, 1997 · Zbl 0997.62510 · doi:10.1201/9781439821862
[17] Wang S, Shao J, Kim J K. An instrument variable approach for identification and estimation with nonignorable nonresponse. Statist Sinica, 2014, 24: 1097-1116 · Zbl 06431822
[18] Zhao J. Identifiability and estimation in statistical models with nonignorable missing data. PhD Thesis. Wilcosin: University of Wilcosin, 2012
[19] Zhu L X, Fang K T. Asymptotics for kernel estimate of sliced inverse regression. Ann Statist, 1996, 24: 1053-1068 · Zbl 0864.62027 · doi:10.1214/aos/1032526955
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.