Hi-
I have a dataset with 3 time points and 3 levels of an independent variable.
I am using a MIXED model to evaluate the following:
(1)Do each of the time points differ from the first data point?
(2a)Are the combination of Conditions 1 and 2 different from Condition 3?
2b Is Condition 1 different from Condition 2?
(3) Is there an interaction of time (with contrasts specified) and treatment
condition (with contrasts specified)?
I can run the contrasts for number 1 and 2 separately (time and condition),
but I cannot seem to run the contrasts for number 3, the interaction model.
When I include all test statements from 1, 2, and 3 in the model, I get a
message that the single level contrasts (1 and 2) are not estimable, but the
interaction terms are given.
Can someone help me? Thanks!
THIS MODEL PROVIDES THE ESTIMATES FOR THE change comparing each time point
to baseline.
MIXED youroutcomehere BY NEWCOND TIME
/FIXED= NEWCOND TIME TIME*NEWCOND
/repeated = TIME | SUBJECT(id) covtype (ar1)
/test 't0 v TIME 1' TIME -1/2 1/2 0 0
/test 't0 v TIME 2' TIME -1/2 0 1/2 0
/PRINT LMATRIX.
*THIS MODEL PROVIDES THE ESTIMATES FOR THE Independent Variable CONTRASTS.
MIXED DEP BY NEWCOND TIME
/FIXED= NEWCOND TIME TIME*NEWCOND
/repeated = TIME | SUBJECT(id) covtype (ar1)
/test 'G1 and G2 v G3' NEWCOND -1/2 -1/2 1
/test 'G1 to G2 ' NEWCOND -1/2 1/2 0
/PRINT LMATRIX.
*THIS MODEL PROVIDES THE ESTIMATES FOR THE INTERACTIONS OF THE CONTRASTS.
MIXED DEP BY NEWCOND TIME
/FIXED= NEWCOND TIME TIME*NEWCOND
/repeated = TIME | SUBJECT(id) covtype (ar1)
/test 't0 v TIME 1' TIME -1/2 1/2 0 0
/test 't0 v TIME 2' TIME -1/2 0 1/2 0
/test 't0 v TIME 3' TIME -1/2 0 0 1/2
/test 'G1 and G2 v G3' NEWCOND -1/2 -1/2 1
/test 'G1 to G2 ' NEWCOND -1/2 1/2 0
/test 't0 v TIME 1 by NEWCOND CONTRAST 1' TIME*NEWCOND 1/4 -1/4 0 0 1/4
-1/4 0 0 -1/2 1/2 0 0
/test 't0 v TIME 1 y NEWCOND CONTRAST 2' TIME*NEWCOND 1/4 -1/4 0 0 -1/4
1/4 0 0 0 0 0 0
/test 't0 v TIME 2 by NEWCOND CONTRAST 1' TIME*NEWCOND 1/4 0 -1/4 0 1/4
0-1/4 0 -1/2 0 1/2 0
/test 't0 v TIME 2 y NEWCOND CONTRAST 2' TIME*NEWCOND 1/4 0 -1/4 0 -1/4
0 1/4 0 0 0 0 0
/test 't0 v TIME 3 by NEWCOND CONTRAST 1' TIME*NEWCOND 1/4 0 0 -1/4 1/4
0 0-1/4 -1/2 0 0 1/2
/test 't0 v TIME 3 y NEWCOND CONTRAST 2' TIME*NEWCOND 1/4 0 0 -1/4
-1/4 0 0 1/4 0 0 0 0
/PRINT LMATRIX.
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
