Review Articles

Impact of sufficient dimension reduction in nonparametric estimation of causal effect

Ying Zhang ,

Department of Statistics, University of Wisconsin-Madison, Madison, WI, USA

Jun Shao ,

Department of Statistics, University of Wisconsin-Madison, Madison, WI, USA; School of Statistics, East China Normal University, Shanghai, People's Republic of China

Menggang Yu ,

Department of Biostatistics and Medical Informatics, University of Wisconsin-Madison, Madison, WI, USA

Lei Wang

Institute of Statistics and LPMC, Nankai University, Tianjin, People's Republic of China

lwangstat@nankai.edu.cn

Pages 89-95 | Received 10 Nov. 2017, Accepted 14 Apr. 2018, Published online: 18 May. 2018,
  • Abstract
  • Full Article
  • References
  • Citations

ABSTRACT

We consider the estimation of causal treatment effect using nonparametric regression or inverse propensity weighting together with sufficient dimension reduction for searching low-dimensional covariate subsets. A special case of this problem is the estimation of a response effect with data having ignorable missing response values. An issue that is not well addressed in the literature is whether the estimation of the low-dimensional covariate subsets by sufficient dimension reduction has an impact on the asymptotic variance of the resulting causal effect estimator. With some incorrect or inaccurate statements, many researchers believe that the estimation of the low-dimensional covariate subsets by sufficient dimension reduction does not affect the asymptotic variance. We rigorously establish a result showing that this is not true unless the low-dimensional covariate subsets include some covariates superfluous for estimation, and including such covariates loses efficiency. Our theory is supplemented by some simulation results.

References

  1. Adichie, J. N. (1974). Rank score comparison of several regression parameters. The Annals of Statistics, 2, 396402. doi: 10.1214/aos/1176342676 [Google Scholar]
  2. Chen, X., Wan, A. T. K., & Zhou, Y. (2015). Efficient quantile regression analysis with missing observations. Journal of the American Statistical Association, 110, 723741. doi: 10.1080/01621459.2014.928219 [Taylor & Francis Online], [Google Scholar]
  3. Cook, R. D., & Weisberg, S. (1991). Discussion of ‘Sliced inverse regression for dimension reduction’. Journal of the American Statistical Association, 86, 328332[Taylor & Francis Online], [Google Scholar]
  4. Deng, J., & Wang, Q. (2017). Dimension reduction estimation for probability density with data missing at random when covariables are present. Journal of Statistical Planning and Inference, 181, 1129. doi: 10.1016/j.jspi.2016.08.007 [Google Scholar]
  5. De Wet, T., & Van Wyk, J. W. J. (1979). Efficiency and robustness of Hogg's adaptive trimmed means. Communications in Statistics: Theory and Methods, 8, 117128. doi: 10.1080/03610927908827743 [Taylor & Francis Online], [Google Scholar]
  6. Gong, G., & Samaniego, F. J. (1981). Pseudo maximum likelihood estimation: Theory and applications. The Annals of Statistics, 9, 861869. doi: 10.1214/aos/1176345526 [Google Scholar]
  7. Hahn, J. (1998). On the role of the propensity score in efficient semiparametric estimation of average treatment effects. Econometrica, 66, 315331. doi: 10.2307/2998560 [Google Scholar]
  8. Hahn, J. (2004). Functional restriction and efficiency in causal inference. Review of Economics and Statistics, 86, 7376. doi: 10.1162/003465304323023688 [Google Scholar]
  9. Hirano, K., Imbens, G. W., & Ridder, G. (2003). Efficient estimation of average treatment effects using the estimated propensity score. Econometrica, 71, 11611189. doi: 10.1111/1468-0262.00442 [Google Scholar]
  10. Hsing, T., & Carroll, R. J. (1992). An asymptotic theory for sliced inverse regression. The Annals of Statistics, 20, 10401061. doi: 10.1214/aos/1176348669 [Google Scholar]
  11. Hu, Z., Follmann, D. A., & Wang, N. (2014). Estimation of mean response via the effective balancing score. Biometrika, 101, 613624. doi: 10.1093/biomet/asu022 [Google Scholar]
  12. Imai, K., & Ratkovic, M. (2014). Covariate balancing propensity score. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 76, 243263. doi: 10.1111/rssb.12027 [Google Scholar]
  13. Imbens, G. W. (2004). Nonparametric estimation of average treatment effects under exogeneity: A review. Review of Economics and Statistics, 86, 429. doi: 10.1162/003465304323023651 [Google Scholar]
  14. Imbens, G. W., Newey, W., & Ridder, G. (2006). Mean-squared-error calculations for average treatment effects (Working Paper). [Google Scholar]
  15. Kang, J. D. Y., & Schafer, J. L. (2007). Demystifying double robustness: A comparison of alternative strategies for estimating a population mean from incomplete data. Statistical Science, 22, 523539. doi: 10.1214/07-STS227 [Google Scholar]
  16. Li, K. C. (1991). Sliced inverse regression for dimension reduction. Journal of the American Statistical Association, 86, 316327. doi: 10.1080/01621459.1991.10475035 [Taylor & Francis Online], [Google Scholar]
  17. Luo, W., Zhu, Y., & Ghosh, D. (2017). On estimating regression-based causal effects using sufficient dimension reduction. Biometrika, 104, 5165[Google Scholar]
  18. Raghavachari, M. (1965). On the efficiency of the normal scores test relative to the F-test. The Annals of Mathematical Statistics, 36, 13061307. doi: 10.1214/aoms/1177700005 [Google Scholar]
  19. Randles, R. H. (1982). On the asymptotic normality of statistics with estimated parameters. The Annals of Statistics, 10, 462474. doi: 10.1214/aos/1176345787 [Google Scholar]
  20. Rosenbaum, P. R., & Rubin, D. B. (1983). The central role of the propensity score in observational studies for causal effects. Biometrika, 70, 4155. doi: 10.1093/biomet/70.1.41 [Google Scholar]
  21. Wang, D., & Chen, S. X. (2009). Empirical likelihood for estimating equations with missing values. The Annals of Statistics, 37, 490517. doi: 10.1214/07-AOS585 [Google Scholar]
  22. Wang, Q. (2007). M-estimators based on inverse probability weighted estimating equations with response missing at random. Communications in Statistics: Theory and Methods, 36, 10911103. doi: 10.1080/03610920601076917 [Taylor & Francis Online], [Google Scholar]
  23. Xia, Y., Tong, H., Li, W. K., & Zhu, L. X. (2002). An adaptive estimation of dimension reduction space. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 64, 363410. doi: 10.1111/1467-9868.03411 [Google Scholar]
  24. Zhu, L. X., & Ng, K. W. (1995). Asymptotics of sliced inverse regression. Statistica Sinica, 5, 727736[Google Scholar]