Statalist The Stata Listserver


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

Re: st: Re: Contrasts and interpretation


From   Joseph Coveney <[email protected]>
To   Statalist <[email protected]>
Subject   Re: st: Re: Contrasts and interpretation
Date   Wed, 13 Dec 2006 23:19:47 +0900

Jaye Brown wrote:

I am trying to analyse a 3-way anova and look at the 3 way interactions
using test and got the following results.

. test _coef[m[1]*s[2]*i[2]] = _coef[m[1]*s[2]*i[3]]
( 1)  m[1]*s[2]*i[2] - m[1]*s[2]*i[3] = 0
      F(  1,   162) =    4.24
           Prob > F =    0.0412

. test _coef[m[2]*s[2]*i[2]] = _coef[m[2]*s[2]*i[3]]
( 1)  m[2]*s[2]*i[2] - m[2]*s[2]*i[3] = 0
      Constraint 1 dropped

I am confused as to why I get the message "Constraint 1 dropped" and does
this mean that I cannot estimate this interaction term?

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

It could be signaling that one or more of the involved coefficients has
been dropped for model identifiability.  Try typing -anova ,
egress-.  -anova- typically drops the highest level of a factor (just like
SAS's PROC GLM and PROC ANOVA), unless you change it with -char
varname[omit] #-.

It's just a guess, but are you sure that your test is really testing the
hypothesis that you intend to test?  -anova , regress- or -anova , symbolic-
would be informative for that, too.  I'd recommend the same thing to Janet
Pol, who posted an analogous contrast earlier this week.

Joseph Coveney

*
*   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