邏輯斯模型有測量誤差時之模擬研究

The simulation study of logistic regression model with measurement errors

張煦(Hsu Chang)

指導教授 ： 黃逸輝

淡江大學/理學院/數學學系碩士班/碩士(2013年)

https://doi.org/10.6846/TKU.2013.00745

摘要

本文探討當邏輯斯模型的自變數有測量誤差時，各種不同的估計方法表現的模擬研究。這些方法包含了傳統的迴歸校正、條件分數及新穎的延伸校正分數函數等等。我們考慮有不同的自變數分佈、樣本大小、迴歸參數值、測量誤差變異數與測量誤差分佈等許多情況，並從估計量的樣本變異數或是標準差的大小來評估不同估計方法的效率。

關鍵字

測量誤差 ； 迴歸校正 ； 條件分數 ； 延伸校正分數

並列摘要

This thesis discusses the performances of different estimation methods in logistic regression model when covariates are subject to measurement errors. The efficiencies of different methods were evaluated through extensively simulation studies. These simulation studies were conducted under different distribution of covariates, samples size, values of regression parameter, the variance associated with measurement errors and the distribution of measurement errors. Conclusions based on the sample standard deviations were derived.

並列關鍵字

measurement errors

參考文獻

Huang, Y. and Wang, C. Y. (2001). Consistent Functional Methods for Logistic Regression with Errors in Covariates. Journal of the American Statistical Association 96, 1469-1482.

Nakamura, T. (1990). Corrected score functions for errors-in-variables models: Methodology and application to generalized linear models. Biometrika 77, 127-137.

Stefanski, L. A. and Carroll, R. J. (1987). Conditional scores and optimal scores for generalized linear measurement error models. Biometrika 74, 703-716.

楊喬閔(2012). Extended corrected-score in generalized linear models. 中華民國淡江大學碩士論文.

Carroll, R. J., Ruppert, D., Stefanski, L. A., and Crainiceanu, C. M. (2006). Measurement Errors in Nonlinear Models: A Modern Perspective. Second Edition. Chapman & Hall, London.