help > Design matrix and contrast for Mixed ANOVA and post hoc simple main effects
Sep 7, 2018  10:09 AM | thayashi423
Design matrix and contrast for Mixed ANOVA and post hoc simple main effects
Dear Prof. Zalesky,

I have 240 scans: 5 times from 48 participants. The data were divided into two groups (24 controls and 24 patients). I will perform a mixed (within-between) ANOVA to identify an interaction between them. If it is significant, I will further perform simple main effects at each time point (e.g., patient vs. control at time 1, patient vs. control at time 2, ...). 

The data are aligned as:
control1-time1
control1-time2
control1-time3
control1-time4
control1-time5
control2-time1
.....
patient1-time1
patient1-time2
patient1-time3
patient1-time4
patient1-time5
patient2-time1
.....,

Then how do the design matrix and contrast look like?  To my understanding, the design matrix would be an attached file. Columns 1-4 indicates Time and Columns 5-8 indicates Interaction. Is it correct? Also, F contrast vector would be [0 0 0 0 1 1 1 1 0 0 0 , ... , 0] for interaction. Then how do you make the contrast vectors for the simple main effects?

Thank you in advance for your help.

Best Regards,

Taku
Attachment: DesignMatrix.mat

Threaded View

TitleAuthorDate
Design matrix and contrast for Mixed ANOVA and post hoc simple main effects
thayashi423 Sep 7, 2018
Andrew Zalesky Sep 9, 2018