Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Interpreting coefficients for a gamma regression with log link (Stata 11)


From   "Weichle, Thomas" <Thomas.Weichle@va.gov>
To   <statalist@hsphsun2.harvard.edu>
Subject   Re: st: Interpreting coefficients for a gamma regression with log link (Stata 11)
Date   Mon, 19 Sep 2011 10:17:53 -0500

Hi Hitesh,
As you know may know, the output being displayed is different depending
on whether you specify the eform option or not.  The resulting model is
the same but the interpretation of the coefficients is slightly
different.  Suppose x is a 1/0 variable.  Then, b =
ln(E(y|x=1))/ln(E(y|x=0)) is the unexponentiated coefficient for x and
e(b) = e^[ln(E(y|x=1))/ln(E(y|x=0))] = e^[ln(E(y|x=1))]/[e^ln(E(y|x=0))]
= E(y|x=1)/E(y|x=0) is the exponentiated coefficient for x.  When
dealing with cost as an outcome, Blough (2000) states that the ratio
directly interpreted from the exposure variable may be preferable to
point estimates (unexponentiated coefficient).

Returning to your specific questions:
1) You are correct in stating that the exponentiated coefficient of
1.031383 represents the arithmetic mean ratio in the dependent variable
between HIV+ and HIV- patients [specifically mean(hiv=1)/mean(hiv=0)].

2) Since 'insurance_3' is the reference category, the exponentiated
coefficient for 'insurance_1' is interpreted as the following ratio:
[mean(insurance_1)/mean(insurance_3)].

Tom Weichle
Math Statistician
Center for Management of Complex Chronic Care (CMC3)
Hines VA Hospital, Bldg 1, C202
708-202-8387 ext. 24261
Thomas.Weichle@va.gov 

*
*   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index