[R] Re: R-help Digest V2 #283
adrian at olsen.ch
Fri Dec 8 10:03:30 CET 2000
R-help Digest wrote:
> Date: Thu, 07 Dec 2000 18:28:09 +0100
> From: Uwe Ligges <ligges at statistik.uni-dortmund.de>
> Subject: Re: [R] Heteroskedasticity in R
> Vincent Leycuras wrote:
> > Hi all,
> > I just discovered R a couple of days ago and I must say it rocks. I've been
> > looking for heteroskedasticity tests and couldn't find any, however.
> > Particularly, I've been told in one of my courses on econometrics of White's
> > method (>< white.test()).
> > The test's statistic is beta / sqrt(W), where W is Var(beta) "à la White",
> > that is the beta(i) matrix is replaced by the residuals matrix.
I think what you mention here is not a test for heteroskedasticity but a robustified (against misspecification, e.g., heteroskedasticity) t-test.
> > Anyway, whatever anybody may haev, I'm interested, else I'll write it this
> > week-end.
> Have a look at the package 'tseries', available at CRAN.
> Uwe Ligges
The White test which Vincent mentions is different from the White Test implemented in tseries, which tests for neglected nonlinearity in the conditional mean, essentially for a regression equation or linear time series model.
One very simple way to test conditional heteroskedasticity (time series setup) is to use the Box.test (package ts) on the squared values of a time series as
Box.test(x^2, lag = 1, type="Ljung")
Adrian Trapletti, Olsen & Associates Ltd., See-
feldstrasse 233, CH-8008 Zürich, Switzerland
Phone: +41 (1) 386 48 48 Fax: +41 (1) 422 22 82
E-mail: adrian at olsen.ch WWW: http://www.olsen.ch
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch
More information about the R-help