常態迴歸分析中, 一個典型的問題就是常被當作是干擾參數的變異數, 其一致的假設是有所違背. 因為如果不理會變異數的結構而將之視為干擾參數, 不但會影響對平均數的估計, 也會導致我們做出錯誤的結論. 然而有很多例子指出, 瞭解變異數的結構形式是非常重要的. 本篇論文的目的就是使用一個有別於常用在處理非齊一變異數的方法: 建立變異數模式. 我們研究另一種型式的變異數模式, 提出一個一般化的形式, 並建議一個圖形化工具幫助我們進行變異數模式的建立.

In practice, an often occured problem in Normal regression analysis is the violation of homogeneous assumption in error variance, which is usually treated as a nuisance parameter. Treating variance structure as a nuisance instead of a central part of the modelling effort not only leads to inefficient estimation of means, but also to misleading conclusions. There are many instances, however, indicate understanding the structure of variability is very important. The purpose of this work is to provide an alternative approach of dealing with non-constant variance: modelling it. We study different forms of modelling the error variance, propose a general formation, and suggest graphical tool to help us modelling the variance function.

Aitkin, M. (1987), "Modelling variance heterogeneity in normal regression
using GLIM," Royal Statistical Society, 36, 332--339.
Atkinson, A. C. (1982), "Regression diagnostics, transformations and
constructed variables (with discussion)," Journal of Royal Statistical
Society, Series B, 44, 1--36.
Boos, D. D., and Brownie, C. (1989), "Bootstrap methods for testing
homogeneity of variances," Technometrics, 31, 69--82.
Box, G. E. P., and Cox, D. R. (1964), "An analysis of transformations,
(with discussion)," Journal of the Royal Statistical Society, series B,
26, 211--246.
Box, G. E. P., and Tidwell, P. W. (1962), "Transformation of the independent
variables." Technometrics, 4, 531--550.
Bradely, E. L. (1973), "The equivalence of maximum likelihood and weighted
least squares estimates in the exponential family," Journal of the American
Statistical Association, 68, 199--200.
Breusch, T. S., and Pagan, A. R. (1979), "A simple test for heteroskedasticity
and random coe cient variation," Econometrica, 47, 1287--1294.
Charnes, A., Frome, E. L., and Yu, P. (1976), "The equivalence of generalized
least squares and maximum likelihood estimates in the exponential family,"
Journal of the American Statistical Association, 71, 169--171.
Conover, W. J., Johnson, M. E., and Johnson, M. M. (1981),"A comparative study
of tests for homogeneity of variances, with applications to the outer
continental shelf bidding data," Technometrics, 23, 351--361.
Cook, R. D., and Weisberg, S. (1983), "Diagnostics for heteroscedasticity in
regression," Biometrika, 70, 1--10.
─ (1999), Applied Regression Including Computing and Graphics, New York: John
Wiley & Sons.
Dobson, A. J. (1990), An Introduction to Generalized Linear Models, London:
Chapman & Hall.
Lee, L. (1980), "Testing adequacy of the weibull and log linear rate models
for a poisson process," Technometrics, 22, 195--199.
Nelder, J. A., and Lee, Y. (1998), "Letters to the editor: "Joint Modeling of
Mean and Dispersion"," Technometrics, 40, 168--171.
Nelder, J. A., and Wedderburn, R. W. M. (1972), "Generalized linear models,"
Journal of Royal Statistical Society, Series A, 135, 370--384.
Rosenberg, P. S., and Gail, M. H. (1991), "Backcalculation of flexible linear
models of the human immunodeficiency virus infection curve," Applied
Statistics, 40, 269--282.
Ryan, T. A., Joiner, B. L., and Ryan, B. F. (1976), The Minitab Student
Handbook, North Scituate, MA: Duxbury Press.
Smyth, G. K. (1989), "Generalized linear models with varying dispersion,"
Journal of Royal Statistical Society, Series B, 51, 47--61.
─ (2002), "An e cient algorithm for reml in heteroscedastic regression,"
Journal of Computational and Graphical Statistics, 2(4), 836--847.
Venables, W. N. (2000), "Exegeses on linear models," the S-PLUS User's
Conference, 1--25, Washington, DC, 8-9th October, 1998.
Williams, E. J. (1959), Regression Analysis, New York: John Wiley & Sons.