Thanks Maarten.
Cheers,
Pek
p/s I also found useful info in your response to someone's question on the
significance of probit estimate.
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Maarten buis
Sent: October-23-08 12:59 AM
To: [email protected]
Subject: RE: st: mfx after xtnbreg and how to compute predicted Y
--- Pek-Hooi Soh <[email protected]> wrote:
> For your first response about a3 being the main effect of x3 when x2
> has it's mean value,
> 1) Is a3 the estimate I obtain with irr option?
If you want to interpret the result as a ratio of risks then you have
to specify the -irr- option.
> 2) Must the estimates for both the single term and the interaction
> term be statistically significant in the results? What if one or the
> other is not, would the interpretation be different? Or it has no
> change as long as the model fit is significant.
Significance and effect size are two different concepts. Significance
is irrelevant if you want to know the size of an effect. The two are
distinct concepts: what is my best guess of the size of the effect,
what is the chance that this effect is just random noice that occured
because I have drawn a random sample.
Hope this helps,
Maarten
-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands
visiting address:
Buitenveldertselaan 3 (Metropolitan), room N515
+31 20 5986715
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
*
* 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/
*
* 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/
