Hi mailinglist,
I have a question regarding DOSS/DODS. In the mailinglist I read that if you do a group analysis, you have to check for interaction between your continuous variable and categorical variable in DODS, and if there's no interaction, you can use DOSS. This however leaves a few questions regarding my analysis:
I want to look at the correlation between cognitive performance and cortical thickness, accounting for gender, age and IQ. In a sense, this is not a 'real' group analysis, since gender is merely a covariate. Do I have to check for interactions, and if so, for which ones?
"Does the cogn perf-thickness correlation differ between male and female?"
"Does the age-thickness correlation differ between male and female?" (age = demeaned)
Last but not least, at which Monte Carlo settings would you usually check for an interaction? As far as I understood from reading the mailinglist, I would think vertexwise threshold at 2, absolute (2 way), clusterwise threshold 1.3. Is this correct? Or would it for example make more sense to test 1 way (positive for cogn perf-thickness, negative for age-thickness)?
Thank you very much,
Iris