| Science Forum Index » Space - Consult Forum » Mixed model with categorical data... |
|
Page 1 of 1 |
|
| Author |
Message |
| Mimi... |
 Posted: Mon Oct 12, 2009 9:04 pm |
|
|
|
Guest
|
I have a dataset including the ID, sex (M/F) and migratory status
(1/0: resident/migrant) of birds. The data were collected over many
years, therefore most individuals occurred several times in the
dataset. I used SAS proc GLIMMIX to fit a model of "migratory status =
sex" with bird ID as the random effect.
My questions are:
1. Is it similar to a logistic regression with categorical independent
variables?
2. What is the proper test statistics to report to show if sex is
relevant to migratory status? The F in the Type III Tests of Fixed
Effects?
3. How is the F derived for categorical variables? What is the error
variance in such case?
Thank you~
Monika |
|
|
| Back to top |
|
|
|
|
