Hi All,
sorry to bother you with what might be to some an 'obvious' question.
I've been helping someone analyse data regarding factors that may
predict/influence whether patients fail to turn up for outpatient
appointments. I basically thought that a logistic regression might be a
sensible approach.
The problem that has occured to me is, that if I did the regression and
(say) it turned out that patients were three times as likely not to
attend on wednesday but that on viewing the data it turned out that
patients were allocated three times as many appointments on a wednesday
(so the increased non attendence is probably due to the increased number
of appointments given on that day) - then where would I go from there?
So in essence I am asking how to (whether you have to) adjust logistic
regression to take into account the fact that the subjects in the
regression may not be distibuted evenly within the categories used to
describe them. I hope that makes sense.
I would be very grateful for any advice you could give me.
Regards
Graham
--
Dr G.S.Clarke
Lecturer in Physiology & Biometery
University of Wales - Bangor
Bangor
Gwynedd LL57 2UW
Tel: 01248 382535
e-mail: [log in to unmask]
|