Thank you Peter for showing me the error.
I did not realize it. Now I have removed that cohort (there was just one
observation!) and checked the numbers for each of the other cohorts. I have
re-run the model and now it seems to make much more sense to me.
I am going to use one specific cohort,
At 08:40 08/05/2012, lincoln wrote:
Thank you Peter for showing me the error.
I did not realize it. Now I have removed that cohort (there was just one
observation!) and checked the numbers for each of the other cohorts. I have
re-run the model and now it seems to make much more sense to me.
I
Hi all,
I can't find the error in the binomial GLM I have done. I want to use that
because there are more than one explanatory variables (all categorical) and
a binary response variable.
This is how my data set looks like:
str(data)
'data.frame': 1004 obs. of 5 variables:
$ site : int 0 0
1. As this is a statistical, rather than an R issue, you would do
better posting on a statistical help site like stats.stackexchange.com
(although some generous soul here may respond).
2. You would also probably do better consulting with a local
statistical resource if available, as it is
Perhaps I haven't explained it that well as I would have liked to.
To me this was an R issue because I didn't understand why the binomial GLM
is getting these results and I believed this was something due to the way I
am implementing it in R, not to the binomial GLM itself.
If I was wrong and
On May 7, 2012, at 19:39 , Bert Gunter wrote:
1. As this is a statistical, rather than an R issue, you would do
better posting on a statistical help site like stats.stackexchange.com
(although some generous soul here may respond).
2. You would also probably do better consulting with a
6 matches
Mail list logo