本篇論文是在探討迴歸分析適用性的問題。針對資料本身，首先我們會討論離群值存在的影響，並提供SAS程式處理資料「如何找出資料中的離群值」的問題。再利用"情況1.和情況2."證明來自任意機率分配的資料，我們都可以把資料當作有限母體去處理，可以在不會違背迴歸的基本假設，使用迴歸分析方法進行模型建立。最後再討論迴歸分析在越過資料範圍以外的估計方法-脊迴歸是否可行。

The problem of the applicability of regression analysis is discussed in this dissertation. To specialize the data set, we first discuss the effect of the outlier existed. Then we use the case1 & case2 to prove the date with any probability can be attended as finite population. We can set up a regression model, and we will not departure the assumption of regression analysis. Finally, we try to discuss ridge regression.

目錄：
1 前言………………………………………………………………… 1
2 如何使用迴歸分析………………………………………………… 1
  2.1. 先天限制……………………………………………………… 1
2.2. 變數選擇對模型的影響……………………………………… 2
2.3. 離群値的鑑定………………………………………………… 3
2.4. 對回歸各種基本假設的檢定………………………………… 4
3 比較兩組不同假設下的母體變異數及殘差值…………………… 5 
3.1. (Case1.)在常態假設下的情況
  (under the normal assumption case)…………………………………6
3.2. (Case2.)討論有限母體的情況
(discuss on finite populations case)………………………………… 7
3.3. 結論( from Case1. & Case2. )……………………………………10
4 對預測不適用的迴歸模型如何處理………………………………11
  4.1. 脊迴歸(ridge regression) ………………………………………12
5 參考文獻……………………………………………………………13
6 附註…………………………………………………………………14
註一 SAS programs for testing outliers………………………………14
  註二 (Case1.)中的細部證明…………………………………………20

[1] RUDOLF J. FREUND and WILSON. (Regression Analysis: Statistical Modeling of a Regression Variable). ACADEMICC PRESS. 1998.
[2] DOUGLASC C. MONTGOMERY and ELIZABETH A. PECK and G. GEOFFREY VINING. (Introduction To Linear Regression Analysis Third Edition). WILEY-INTERSCIENCE. 2001.
[3] KUTNER NACHTSHEIM NETER (Applied Linear Regression Models) MC GRAW HILL. 2004.
[4] Hoerl, A. E., and Kennard, R. W.(1970):Ridge regression: Applications to nonorthogonal problems, Technometrics, 12, 69-82.
[5] Kunugi, T., Tamura, T., and Naitio, T. (1961): New acetylene process uses hydrogen dilution, Chemical Engineering Progress, 57, 43-49.
[6] Marquardt, D. W. (1974): Discussion of “The Fitting of Power Series, Meaning Polynomials, Illustrated on Band-Spectroscpic Data,”by A. E. Beaton and J. W. Tukey, Technometrics, 16, 189-192.