[Date Index] [Thread Index] [Author Index]

Mma & ANOVA : package for calculating sums of squares????

• Subject: [mg2229] Mma & ANOVA : package for calculating sums of squares????
• From: purswell at netplus.net (J.P. and/or Valerie Purswell)
• Date: Tue, 17 Oct 1995 06:35:06 GMT
• Approved: usenet@wri.com
• Distribution: local
• Newsgroups: wri.mathgroup
• Organization: Internet Access Plus
• Sender: daemon at wri.com ( )

I am trying to do a four-factor repeated measures ANOVA with Mma.  
neither one will give me Sums of Squares of Mean Squares for my main effects 
and interactions.  I have been (rather tediously) collecting data within 
cells, doing the summing and squaring of Sums.  Before I spend another several 
hours on the analysis, I was wondering if anyone else has figured out how to 
get Sums of Squares and Mean Squares from the Regress or Designed Regress 
packages.  I can compose my own F ratios.  

As the design is a repeated-measures or within - subjects design, the 
appropriate error terms are different from the between-subjects or one 
measurement per experimental unit, so I really need the Sums of Squares for 
each effect.   

Alternatively, has anyone written a package that does this?  I've looked 
through mathsource but did not find anything....