[R-sig-ME] Interpretation of lmer's sigma for binomial data

[David Duffy]

On Mon, 2 Aug 2010, Rolf Turner wrote:

>
> I would like to second the request for enlightenment on this issue.
> In particular, how does this fit with the recent posting from Jarrod
> Hadfield:
>
>> Over-dispersion does not occur with a binary response variable so you
>> don't need to test for it.

> So what is the quasibinomial family ***doing***, then?

The only trouble with this differentiation of Bernoulli data from 
Binomial data is that we usually fit both types of data using the same 
model.  Which is say, dispersion problems are seen once you are 
trying to model the effects of a covariate, which is implicitly present 
for the binomial setup.

Cheers, David Duffy.
-- 
| David Duffy (MBBS PhD)                                         ,-_|\
| email: davidD at qimr.edu.au  ph: INT+61+7+3362-0217 fax: -0101  /     *
| Epidemiology Unit, Queensland Institute of Medical Research   \_,-._/
| 300 Herston Rd, Brisbane, Queensland 4029, Australia  GPG 4D0B994A v