st: Re: Anovalator output

["Joseph Coveney" <jcoveney@bigplanet.com>]

Martyn Sherriff wrote:

I am running xtmixed on a repeated measures design that is unbalanced,
there are 4 levels of treatment and 7 levels of time. I am then using
anovalator (from http://www.ats.ucla.edu/stat/stata/ado/analysis) to
investigate the main and simple effects.
xtmixed f trt##time || id: , reml

. anovalator time, main
anovalator main-effect for time
chi2(1) = 134.57337   p-value = 4.093e-31

. anovalator trt, main
too few variables specified
r(102);

Can anybody suggest why I get this error message for 'trt'.

--------------------------------------------------------------------------------

How imbalanced are things?  Singletons?  -anovalator- seems to work with 15% 
missing values (below).  Perhaps you could -set trace on-, see just what command
inside balks, and look further from there.

Joseph Coveney

. clear *

. set more off

. set seed `=date("2011-12-08", "YMD")'

. set obs 200
obs was 0, now 200

. generate int pid = _n

. generate double u = rnormal()

. 
. generate byte group = mod(_n, 4)

. 
. forvalues time = 1/7 {
  2.         generate double response`time' = u + ///
>                 `time' / 7 * 0.1 + rnormal()
  3. }

. quietly reshape long response, i(pid) j(time)

. 
. drop if runiform() < 0.15
(198 observations deleted)

. 
. quietly xtmixed response i.group##i.time || pid: , ///
>         reml nolrtest nostderr nolog

. anovalator time, main

anovalator main-effect for time  
chi2(6) = 7.3372743   p-value = .29077902

. anovalator group, main

anovalator main-effect for group  
chi2(3) = .20230331   p-value = .97721661

. exit

end of do-file


*
*   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/