RE: st: anova account for the same individuals

["HUANG, LILING" <HUANGL@mailbox.sc.edu>]

David,

I have 11.2 version. I guess this version does not support contrast commands. I cannot run the contrast and margin commands on my data, can I? 

Does the tukeyhsd command account for whether the observations are the same people, like with a paired t-tests? The tukeyhsd command is not official Stata, so it is not good to report its results in publications, right?
I did the paired t-tests, ttest score_ad1=score_ad2 for pairwise comparions, after I ran the two-way ANOVA omnibus test you suggested. What are the graphs that I should use to be okay to report my results in publication?

How about the xtmixed command? Can I report its results alone without the omnibus test, or combine it with the omnibus test?
xtmixed score ad1_ad2 ||id:, cov(identity)
ad1_ad2 is a dummy variable to specify the pair to be tested.

Again, I really appreciate your help with my questions. I do not have any other statistic support at the moment.

Liling
________________________________________
From: owner-statalist@hsphsun2.harvard.edu [owner-statalist@hsphsun2.harvard.edu] on behalf of Airey, David C [david.airey@vanderbilt.edu]
Sent: Tuesday, November 19, 2013 9:02 PM
To: <statalist@hsphsun2.harvard.edu>
Subject: Re: st: anova account for the same individuals

.

If you choose to use ANOVA for the omnibus test, you can follow up with paired t-tests as the posthoc test for pairwise comparisons of ads.

Be sure to say what version of Stata you have, because this allows advice on use of margins or contrast commands. You mention a tukeyhsd command that is not official Stata.

While you could also investigate other commands like manova, xtgee, xtmixed, or xtologit, if you don't have a lot of statistical support near you, careful use of graphs combined with anova and posthoc paired t-tests is probably OK.

> Dear David,
>
> Thank you very much for your response to my inquiry. For this two-way analysis of variance, can I run pairwise comparisons to test the difference in rating scores between two ads, e.g., tukeyhsd? How do I specify this in the post-hoc test?
> The scale of values for the rating scores is a 5-point Likert scale, ranging from 1 to 5. Indeed, this scale was created from 5 variables with values ranging from 1-5.  If the values are too few to run ANOVA, which method would be appropriate, xtmixed?
> Thanks,
>
> Liling




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