RE: st: repeated measures ANOVA to MANOVA - revisit [on behalf of Rob Ploutz-Snyder]

["Airey, David C" <david.airey@vanderbilt.edu>]

.

Don't expect the p-values between repeated measures ANOVA and MANOVA to be
the same.

There is nothing wrong with Stata's repeated measures ANOVA implementation.

There are different ways of parameterizing such models with ANOVA, as Phil Ender has
said on this list at different times, and you can get each depending on what you use
for your errors.

What may be wrong with RMANOVA is the approach itself, not the implementation in Stata.

The assumptions about the variance covariance matrix are more strict with ANOVA, 
(compound symmetry) which is why some prefer MANOVA which assumes much less. 
But MANOVA can be less powerful, and in some cases you cannot perform MANOVA 
when you can perform RMANOVA. Modern approaches to repeated measures data include
GEE and mixed models. ANOVA is strict about the correlation
structure, MANOVA is not, mixed models allow you model the covariance
structure, and GEE treats it like a nuisance.

When appropriate, response feature analysis can be very clear to interpret. 
For example, if you can change repeated measures to areas under the curve you 
can just analyze one response per subject. This may be useful when the same
measure is taken over time, but not when subjects receive all treatments.

-Dave





> Thank you Rob.
> 
> This entire thing started because we submitted a manuscript using repeated measures ANOVA and one of the reviewers asked to redo the analysis using MANOVA which he/she feels is better than to adjust the df using Huynh-Feldt or some other method. Although I tend to agree with the premise, I am concerned and confused with the large difference in the p-values that the ANOVA and ANOVA report.
> 
> I will like to stay away from -xtmixed- mainly because this is an "accept with minor revisions" paper and I do not want to confuse the reviewers.
> 
> Ricardo.   




*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/