re: st: margins after split-plot anova

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

.

Read the anova manual entry for some examples using the margins command. You probably have to use the -,within()- option to margins to get what you want. In a split plot, subjects are nested in the between subject factor (group), and crossed with the within subject factor (label). See example 13 in the -anova- manual entry.


> Dear Stata users, 
> unfortunately I can’t get margins work after anova. 
> 
> My model includes a between subject factor (group), a continuous covariate (IQ), and a within subject factor (label): 
> anova depvar c.IQ  group / id|group label  label#group  label#c.IQ ,  repeated (label) 
> 
> The command: margins group
> gives no results (not estimable)
> 
> The command (suggested elsewhere): margins group, asbalanced emptycells(reweight)
> gives no results (not estimable)
> 
> The command: margins group, asbalanced emptycells(reweight) noestimcheck
> gives results, but they are clearly unreliable; I mean, they are clearly wrong given the empirical data that I know. 
> 
> The command: margins group,  noestimcheck
> gives results, but, as above, results are wrong (and slightly different from the previous command). 
> 
> 
> If I exclude the continuous covariate (IQ) from the anova model, margins works correctly. But I’d like to keep the continuous covariate in the model. 
> 
> Any help would be appreciated. 
> Thank you.
> Luca


*
*   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/