Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

re: st: Re: Anova and contrasts


From   David Airey <[email protected]>
To   [email protected]
Subject   re: st: Re: Anova and contrasts
Date   Mon, 11 Dec 2006 10:02:56 -0600

I am interested in looking at the system*material interactions in more detail. I have carried out "all possible" comparisons using:
.. test _coef[system[1]*material[1]]=_coef[system[1]*material[2]]
etc.
It occurs to me that I should setup a contrast matrix test this using the mtestand allow for multiple testing. The problem is that the design matrix has 54 columns. Do I need to set up this matrix for the 4 comparisons that I need or is there a more efficient (or simpler) approach.
Many thanks.
When you get a significant interaction, there are two ways to follow it up. One is with simple main effects, the other is with interaction contrasts. Seems you are doing the latter, and you see the prospect of typing in a big matrix annoying? Is that the question? Maybe there is a quicker way to type that in? I'd be interested in the answer too. You might also try the look at simple main effects with the command sme (findit sme). An example is at UCLA:

http://www.ats.ucla.edu/stat/Stata/faq/sme.htm

Another possibility, if you have done all comparisons as you said, is to pass those statistics or p values to a multiple test correction command like smileplot (findit smileplot).

-Dave
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/




© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index