re: st: tukey/dunnett procedures

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

> hi,
> 
> I am looking for any programs, or advice, on doing post-hoc Tukey and
> Dunnett procedures in Stata following a nested ANOVA command
> 
> I am familiar with Gleason's -prcomp- command but it cannot handle nesting
> 
> thanks,
> 
> Rich



I think if the nesting is because multiple measures were made to improve the measurements rather than estimate variance components, I would just mean them and use fixed effects oneway ANOVA if I wanted to use Dunnett's procedure for comparisons with a control, or Tukey's procedure for all pairwise comparisons. There are so many of these procedures (all seem oriented to fixed effects ANOVA), that I have also just used the full model error in pairwise comparisons and adjusted by Holm's sequential Bonf. procedure.
*
*   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/