[R] geepack(geese) vs. glmmPQL

[Thomas Lumley]

On Thu, 13 Mar 2003, Dieter Menne wrote:

> Dear R-List,
>
> I have a data set (available on request) of observations from
> a 2x2 conditions psychological experiment, with count of
> correct/wrong decisions, and vp as the subject id.
<snip>
>  $ decision: Factor w/ 2 levels "corr","wrong": 1 2 1 2 1 2 1 2 1 2 ...

This is your problem.  geese() requires binary data to be 0,1, not
"corr","wrong"


	-thomas