Message-ID: <1336462311928-4616774.post@n4.nabble.com>
Date: 2012-05-08T07:31:51Z
From: Simone Santoro
Subject: Binomial GLM, chisq.test, or?
In-Reply-To: <CANdJ3dXT1L9MT81+5B4u+WNC=SoV_eKXuKt1E6=W1ZSJh9yBMQ@mail.gmail.com>
Hi Tal,
Thanks for replying.
(1) I am going to use cohort as a factor and (2) no, there are no strong
correlation between "cohort" and the other predictors.
I am using a binomial GLM and the lack of significance of "cohort" seems it
was due to one of the 11 levels (the base level) of this factor to be very
scarcely represented, in other words there was just one individual of cohort
1996.
I did not realize it before
http://r.789695.n4.nabble.com/Can-t-find-the-error-in-a-Binomial-GLM-I-am-doing-please-help-tc4615340.html
someone made me point it out.
--
View this message in context: http://r.789695.n4.nabble.com/Binomial-GLM-chisq-test-or-tp4608941p4616774.html
Sent from the R help mailing list archive at Nabble.com.