Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

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

st: gllamm and pweights with link(log)

From   "Hillel Alpert" <>
To   <>
Subject   st: gllamm and pweights with link(log)
Date   Tue, 16 Nov 2010 13:53:51 -0500

If Dr. Rabe-Hesketh sees this list, I shall be grateful if she could kindly address. Is it "small bug", or something else needed to run this model in gllamm?

Hillel Alpert

On Mon, Nov 15, 2010 at 3:03 PM, Hillel Alpert  wrote:
> I appreciate your interest in helping.
> "Does not work" means the following error message, which returns with link(log) specified and not with link(logit).
> pweight not allowed
> r(101);

I suspect you uncovered a small bug that comes up with this specific
combination of options. You'd be best off figuring this out with
Sophia Rabe-Hesketh, the author of GLLAMM.

Stas Kolenikov, also found at 
Small print: I use this email account for mailing lists only.

As a follow-up to this question, I understand now that the link(log) function provides the risk ratios. However, the gllamm command does not accept pweights with this link function. Can anyone advise how probability weights could be incorporated into such an analysis?

Could someone advise please, does GLLAMM provide relative risk ratios as output?
With the following command, for example, the exponentiated output is given. Is it the odds ratio, and if so, what is the syntax for relative risk ratio?

gllamm depvar var1-var6, i(id) family(binomial) link(logit) pweight(weight) nip(8) adapt eform

Thank you for any information.

Hillel Alpert

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index