Date:         Mon, 18 Dec 2006 18:07:22 -0600
Reply-To:     Robin High <robinh@UNLSERVE.UNL.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Robin High <robinh@UNLSERVE.UNL.EDU>
Subject:      Re: A q about how to do a hotelling t square test
Comments: To: Minze Su <slhappyls@GMAIL.COM>
In-Reply-To:  <200612182236.kBIK5Mls016021@mailgw.cc.uga.edu>
Content-Type: TEXT/PLAIN; charset=US-ASCII

>   I could not find a way to do hotelling t square test in SAS. Can anyone
> help?
>

Hi Minze,

As Paul suggested, you can do equivalent tests with PROC GLM in
multivariate mode. To convert that output to T^2 itself do the following:

for one group, to test if two or more means equal 0:

  T_2 = (n1-1)*((1/(Wilks Lambda) ) - 1)

for two groups, to test if the vectors of means are equal:

  T_2 = (n1 + n2 -2) * ((1/ Wilks_Lambda)-1 )

where n1, n2 are the sample sizes of respective groups.

Rencher's book on Multivariate Analysis (Chapter 5) has formulas for other
multivariate stats to T^2.

Robin High