I have a problem interpreting the output of the mixed model procedure in SPSS
. I found a couple of threads dealing with similar problems, but none helped me solve it.
I have a 2×2 repeated measures crossover design with two fixed factors (medication (A/B) and genotype (A/B)) and a random factor (timepoint (1/2)).
As I understand it, the output for the fixed effects is the general influence of the factors over all subgroups and the parameter estimates test the differences between levels. As I only have two levels for all my factors of interest the output should provide information about the difference between all levels from one that is used as 'baseline'.
This is the output:
Fixed effects Typ IIIa
source df df F p
Const 1 2,798 1753 0,000
gen*med 1 30,945 0,988 0,328
genotype 1 31,329 4,564 0,041
medication 1 30,953 12,886 0,001
a. Dependent variable: RT.
Parameter estimates
Parameter est sd df t p conf. int. 95 %
constant 385,695463 10,256203 4,450 37,606 0,000 358,320557 413,070370
[med=B] * [gen=A] -6,220872 8,462102 40,021 -0,735 0,467 -23,323140 10,881395
[med=A] * [gen=A] -19,875937 8,460116 39,989 -2,349 0,024 -36,974613 -2,777261
[med=B] * [gen=B] 7,742406 4,340784 30,949 1,784 0,084 -1,111266 16,596079
[med=A] * [gen=B] 0b 0 . . . . .
[gen=A] 0b 0 . . . . .
[gen=B] 0b 0 . . . . .
[med=B] 0b 0 . . . . .
[med=A] 0b 0 . . . . .
The questions are: is it correct to interpret this output as that there is a difference between medication(A) in genotype(A) compared to medication(A) in genotype(B)?
If that is the case then I do not understand how to compare the effect of medication within genotype(A). Am I doing something wrong?
I assume this output is correct for multiple comparisons. Is that correct?
The code is:
MIXED RT BY genotype medication timepoint
/FIXED=genotype*medication genotype medication | SSTYPE(3)
/METHOD=ML
/PRINT=SOLUTION
/RANDOM=timepoint | COVTYPE(VC)
/REPEATED=medication*timepoint | SUBJECT(id) COVTYPE(CS).
I would really appreciate if someone could help here 🙂
Best Answer
(I don't know if it is proper to answer old questions, especially with a shaky answer, but here goes.)
I'm only learning mixed models myself, but I think you are not doing anything wrong. The parameter estimates are simply calculated in a way that one level of a categorical variable becomes the reference point to which every other level is compared. I think there is a way to change the reference point (at least change from last to first, not sure if you can pick it freely), but in case you want to do multiple comparisons between levels, the answer is not to change the reference point, but to do separate pairwise comparisons. In your case, it would mean adding the following row to the syntax:
(or perhaps COMPARE(medication), if you prefer them shown in another order).
With EMMEANS you see the estimates in a more natural form (and not in form of effects relative to intercept) and can do pairwise comparisons between individual levels. However, if there are many comparisons, you should remember to do some kind of correction, possibly with a false discovery rate procedure (http://en.wikipedia.org/wiki/False_discovery_rate).