[R] naive variance in GEE

[Qiong Yang]

Thanks Thomas. Here is a further question I asked and the answer from 
Prof Ripley. This agrees with your previous response.

Is it true that the same algorithm was used in calculation of GLM S.E. 
and GEE naive S.E.,
and the only difference is the stopping criteria?

Professor Ripley: No, but it is one of the more important ones.
-----------------------------------------

Thanks,
Qiong

Thomas Lumley wrote:
>
> Sorry, I misread your message. Prof Ripley is right, as usual -- the 
> estimates use different stopping criteria and so are just numerically 
> different.
>
>     -thomas
>
> On Tue, 9 Sep 2008, Thomas Lumley wrote:
>
>> On Mon, 8 Sep 2008, Qiong Yang wrote:
>>
>>> Hi,
>>>
>>> The standard error from logistic regression is slightly different 
>>> from the naive SE from GEE under independence working correlation 
>>> structure.
>>
>> Yes
>>
>>> Shouldn't they be identical? Anyone has insight about this?
>>
>> No, they shouldn't. They are different estimators of the same 
>> quantity, like the mean and median of a symmetric distribution.
>>
>>     -thomas
>>
>>
>>
>>
>>> Thanks,
>>> Qiong
>>>
>>> a<-rbinom(1000,1)
>>> b<-rbinom(1000,2,0.1)
>>> c<-rbinom(1000,10,0.5)
>>> summary(gee(a~b, id=c,family="binomial",corstr="independence"))$coef
>>> summary(glm(a~b,family="binomial"))
>>>
>>> ______________________________________________
>>> R-help at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-help
>>> PLEASE do read the posting guide 
>>> http://www.R-project.org/posting-guide.html
>>> and provide commented, minimal, self-contained, reproducible code.
>>>
>>
>> Thomas Lumley            Assoc. Professor, Biostatistics
>> tlumley at u.washington.edu    University of Washington, Seattle
>>
>
> Thomas Lumley            Assoc. Professor, Biostatistics
> tlumley at u.washington.edu    University of Washington, Seattle