Re: heteroscedastic t-distribution

Bjoern Saentis wrote:

> All,
>
> Is there an extension of the heteroscedastic t-distribution over a
> multi-variate normal distribution? I want to compare the means of
> different n-variate normal distributions, each of which might come
> with a different covariance. Any reference would be great.
>
> Thanks,
> Bjoern
I think you are better off posting this question to the stat 
newsgroups.  I have done so with this reply.

--
Stephen J. Herschkorn                        [EMAIL PROTECTED]