Date:         Wed, 9 Apr 2003 11:22:04 -0400
Reply-To:     Jay Weedon <jweedon@EARTHLINK.NET>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Jay Weedon <jweedon@EARTHLINK.NET>
Organization: http://extra.newsguy.com
Subject:      Re: Wald Test in SAS
Content-Type: text/plain; charset=us-ascii

On Wed, 9 Apr 2003 08:25:45 +0200, "L. Kaminski" <leonneon@web.de>
wrote:

>Hi,
>
>> Well...I'm not sure what role the different observations play, but in
>> order to carry out a Wald test you need the variance-covariance matrix of
>> the parameter estimates.  You don't indicate that you have this in your
>> email.
>
>The parameters a to e are estimated by the GMM Modell. So I think I can get
>the variance-covariance matrix. Can you tell me, where I find a sample, what
>the SAS Programm would look like to do the Wald Test?

It's a straightforward application of matrix algebra, so the simplest
approach would be to use SAS/IML.

The formula is as follows:

If e is a column vector of estimates, and V is its estimated
covariance matrix, then the Wald statistic is given by

        W = Transpose(e) x Inverse(V) x e.

You can use the probchi() function or whatever to obtain a p-value.

JW