LSMEANS produces the CONDITIONAL adjusted mean (conditional on the x-bars).
Marginal effects at the mean are equivalent. In my opinion, conditional
means should only be used with linear (mean based) models. For a more
complete discussion, see (as usual) Korn, E. & Graubard, B. "Marginal
Predictions for Survey Data" Biometrics, about June 1999. I don't have the
article handy, so I might be a bit off on the title. You can ignore the
survey design issues if you don't have survey data. The rest stays the
same.
Bryan Sayer
Statistician, SSS Inc.
[email protected]
-----Original Message-----
From: Cecile DELHUMEAU [mailto:[email protected]]
Sent: Monday, August 26, 2002 8:18 AM
To: [email protected]; [email protected]
Subject: st: Equivalent in Stata for lsmean command from SAS
> You would get better advice, I bet, if you told us exactly what LSMEANS
does.
LSMEANS produces the adjusted mean for each class of a variable of interest.
Exemple :
proc glm;
class age Cas sexec;
model vitA = age Cas sexec;
lsmeans Cas / stderr;
run;
In this case, lsmean give the mean ajusted by sexe and age of vitamine A for
each class of the variable Cas : case and control.
Equivalent in STATA : ???
Thank's for your help !
Cecile
*
* 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/
*
* 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/
