[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: poisson vs. ologit for pinprick data
I've been off this list for a while and am greatly
enjoying my return to this excellent resource and
I'm seeking advise for the appropriate model for
pinprick data scaled from 1 to 5 where 1 represents
the most sensitive and 5 the least to the pinprick
stimulus. I need to model this as a function of
several dependent variables. The pinprick data are
distributed something like Poisson, but the variance
is considerably less than the mean (mean=1.28,
var=0.52), and it does not seem to be a real "count"
type of outcome. Could a poisson model be appropriate
here? Or an ordered logit? Any other suggestions?
Thanks very much.
Do you Yahoo!?
New Yahoo! Photos - easier uploading and sharing.
* For searches and help try: