轉換到正確的模型方法探討

In Study Transformations to Correct Model Inadequacies

林家民(Chia-Min Lin)

指導教授 ： 王國徵

淡江大學/理學院/數學學系碩士班/碩士(2012年)

https://doi.org/10.6846/TKU.2012.01201

摘要

本論文是在討論迴歸分析中常用的一個工具-模型診斷，它常被拿來檢視是否符合迴歸基本假設：常態分配，變異數的一致性以及資料的獨立性。若不滿足變異數一致性時，依照以往大多是看殘差圖決定做什麼動作，是否有個檢定統計量來做檢查，以及是否能找到一個方法來解決。本篇論文將提供數個SAS MACRO程式用作分析，並討論在實務上所會面對的問題以及可能的解決方案。

關鍵字

模型轉換

並列摘要

This paper is to discuss a tool usually used in regression-model of diagnosis, it often used to view whether the return to the basic assumptions; normal distribution, constant variance, and data independence. If it not meets the constant variance in accordance with past, most of all figure tell a story, whether a test statistic to the inspection, and whether they could find a way to solve the problem, so this paper will use several SAS MACRO programs to analyze, and discuss the problems faced in practice as well as possible solutions

並列關鍵字

Transformation

參考文獻

i. Bartlett, M. S. (1937). "Properties of sufficiency and statistical tests". Proceedings of the Royal Statistical Society, Series A 160, 268–282 JSTOR 96803

ii. Bickel, Peter J. (1978). "Using Residuals Robustly I: Tests for Heteroscedasticity", Nonlinearity. Ann. Statist. 6, 266-2912

iii. Brown, Morton B. and Forsythe, Alan B. (1974), "Robust Tests for Equality of Variances", Journal of the American Statistical Association, 69, 364–367

v. Box, G. E. P. & Tidwell, Paul W. (1962). "Transformation of the Independent Variables, " Technometrics, Vol. 4, No. 4 (Nov., 1962), pp. 531-550

vi. Cook, R. Dennis & Wesberg, Sanford. "Diagnostics for Heteroscedasticity in Regression"Biometrika", Vol. 70, No. 1 ( Apr. 1983), 1 – 10.