×

zbMATH — the first resource for mathematics

Multicovariate-adjusted regression models. (English) Zbl 1431.62167
Summary: We introduce multicovariate-adjusted regression (MCAR), an adjustment method for regression analysis, where both the response \((Y)\) and predictors \((X_{1}, \ldots , X_p)\) are not directly observed. The available data have been contaminated by unknown functions of a set of observable distorting covariates, \(Z_{1}, \ldots , Z_s\), in a multiplicative fashion. The proposed method substantially extends the current contaminated regression modelling capability, by allowing for multiple distorting covariate effects. MCAR is a flexible generalisation of the recently proposed covariate-adjusted regression method, an effective adjustment method in the presence of a single covariate, \(Z\). For MCAR estimation, we establish a connection between the MCAR models and adaptive varying coefficient models. This connection leads to an adaptation of a hybrid backfitting estimation algorithm. Extensive simulations are used to study the performance and limitations of the proposed iterative estimation algorithm. In particular, the bias and mean square error of the proposed MCAR estimators are examined, relative to a baseline and a consistent benchmark estimator. The method is also illustrated with a Pima Indian diabetes data set, where the response and predictors are potentially contaminated by body mass index and triceps skin fold thickness. Both distorting covariates measure aspects of obesity, an important risk factor in type 2 diabetes.

MSC:
62G08 Nonparametric regression and quantile regression
62G07 Density estimation
62G20 Asymptotic properties of nonparametric inference
62P10 Applications of statistics to biology and medical sciences; meta analysis
PDF BibTeX XML Cite
Full Text: DOI
References:
[1] Kaysen G. A., Kidney International 61 pp 2240– (2003) · doi:10.1046/j.1523-1755.2002.00076.x
[2] Şentürk D., Biometrika 92 pp 75– (2005) · Zbl 1068.62082 · doi:10.1093/biomet/92.1.75
[3] Ichimura H., Journal of Econometrics 58 pp 71– (1993) · Zbl 0816.62079 · doi:10.1016/0304-4076(93)90114-K
[4] Fan J., Journal of the Royal Statistical Society, Series B 66 pp 57– (2003) · Zbl 1063.62054 · doi:10.1111/1467-9868.00372
[5] Cleveland W. S., Statistical Models in S pp 309– (1991)
[6] Hastie T., Journal of the Royal Statistical Society, Series B 55 pp 757– (1993)
[7] Ramsay J. O., The Analysis of Functional Data (1997) · Zbl 0882.62002 · doi:10.1007/978-1-4757-7107-7
[8] Nicholls D. F., Lecture Notes in Statistics 11 (1982) · doi:10.1007/978-1-4684-6273-9
[9] Chen R., Journal of the American Statistical Association 88 pp 298– (1993)
[10] Hoover D. R., Biometrika 85 pp 809– (1998) · Zbl 0921.62045 · doi:10.1093/biomet/85.4.809
[11] Chiang C., Journal of the American Statistical Association 96 pp 605– (2001) · Zbl 1018.62034 · doi:10.1198/016214501753168280
[12] Cai Z., Journal of the American Statistical Association 95 pp 888– (2000) · doi:10.1080/01621459.2000.10474280
[13] Carroll R. J., Journal of the American Statistical Association 92 pp 477– (1997) · doi:10.1080/01621459.1997.10474001
[14] Xia Y., Journal of the American Statistical Association 94 pp 1275– (1999) · doi:10.1080/01621459.1999.10473880
[15] Şent”urk D., Computational statistics and Data Analysis 50 pp 3294– (2006) · Zbl 1445.62083 · doi:10.1016/j.csda.2005.06.001
[16] Fan J., Local Polynomial Modelling and its Applications (1996) · Zbl 0873.62037
[17] Wahba G., Applications of Statistics pp 507– (1977)
[18] Craven P., Numerical Mathematics 31 pp 377– (1979) · Zbl 0377.65007 · doi:10.1007/BF01404567
[19] Bickel P. J., Journal of the American Statistical Association 70 pp 428– (1975) · doi:10.1080/01621459.1975.10479884
[20] American Diabetes Association, Diabetes Care 23 pp 381– (2000) · doi:10.2337/diacare.23.3.381
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. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.