| Science Forum Index » Statistics - Education Forum » Random effects?... |
|
Page 1 of 1 |
|
| Author |
Message |
| Penfold... |
 Posted: Tue Sep 29, 2009 12:36 am |
|
|
|
Guest
|
Hi
I have some variables measured on group level and some on the
individual level (individuals from the same groups) for 3 different
treatments. Each treatment was repeted 4 times in a random order. The
plan is to make a mixed model of the data to see if there is an effect
of treatment. However, I am in doubt what my random effects in the
individual measurement case should be in order to get the level of
degrees of freedom correct since individuals within a group might not
be independent.
1) 8 groups x 3 treatments x 4 repetitions = 96 df
Random effects: group and group x date
2) 4 individuals x 8 groups x 3 treatments x 4 repetitions = 384 df
Random effects: individual x group and group x date or individual x
group x date???
Thanks a lot |
|
|
| Back to top |
|
|
|
|
