Hi all,
I have a repeated measures design with two groups (drug and placebo) and two time points (baseline and follow-up). The groups are not very well matched so I'd like to control for age and gender. The data was also acquired across three sites which I'll need to control for. I do not have any drop outs.
Essentially, I want to know if there is a significant effect of drug on cortical thickness.
I've tried this out in qdec, but because I have site (with three levels) qdec isn't appropriate, so I'm trying to set up my own fsgd and contrast matrix. Because I have already created the percent change and symmetrized percent change files (as part of qdec prep) I thought I might be able to simply run a t-test on those. Which would answer the question...'is there a significant difference in percent change for the drug and placebo groups'. Does this seem like a sensible approach?
If the above does seem sensible, I am not quite sure how to set up my contrast matrix. As I have three categorical variables to include (Treatment, Gender, Site) and only one continuous variable (age). This would leave me with 12 classes (DrugFemaleSite1, DrugMaleSite1, DrugFemaleSite2, DrugMaleSite2 etc etc) and one variable.
Would it look something like this? Or is this absurd?
0.333 0.333 0.333 0.333 0.333 0.333 -0.333 -0.333 -0.333 -0.333 -0.333 -0.333 0 0 0 0 0 0 0 0 0 0 0 0
Thanks in advance,
Amanda