home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.math.stat
- Path: sparky!uunet!haven.umd.edu!darwin.sura.net!wupost!cs.utexas.edu!uwm.edu!linac!uchinews!spssig.spss.com!nichols
- From: nichols@spss.com (David Nichols)
- Subject: Re: Greenhouse-Geisser Correction Algorithm Needed
- Message-ID: <Jul24.210124.41219@spss.com>
- Date: Fri, 24 Jul 1992 20:49:40 CUT
- Nntp-Posting-Host: spssrs2.spss.com
- Organization: SPSS Inc.
- Lines: 23
-
- The formula for the Greenhouse-Geiser epsilon for a within subjects effect
- (main effect or interaction), from the MANOVA chapter in the SPSS Statistical
- Algorithms, 2nd ed. is:
-
- 2
- (trace(SSE))
- epsilon = -------------------
- 2
- k * trace((SSE) )
-
- where k = degrees of freedom for the factor or interaction
- SSE = Error SSCP matrix for that effect
-
- In other words, on top you're getting the trace and then squaring it. On
- the bottom you're squaring the SSE matrix, then taking the trace and
- multiplying it by k. I checked this out running an example in MANOVA,
- and the formula gives the same results as MANOVA does.
-
- --
- David Nichols Statistical Support Specialist SPSS, Inc.
- Phone: (312) 329-3684 Internet: nichols@spss.com Fax: (312) 329-3657
- *******************************************************************************
- Any correlation between my views and those of SPSS is strictly due to chance.
-