Dear all,
I encountered several questions when using freesurfer. Although I searched carefully in this mailing list, I am still not very clear about some issues. My issues are described below:
Based on a bunch of information, I know that before describing my results, I need to check whether there is an interaction effects between group and age. In order to do it, I demean my age variable and add it as a continuous covariate in Qdec. My discrete variable is group. Then I run the analysis on cortical thickness, producing several results. But I am not very sure about which of them described the group*age interaction effects in my data so that I could check the interaction results. That is why I need your help with confirming it. The name of several results are listed here:
1. does the average thickness differ from zero? (I don't think this is interacton effect) 2. does the correlation between thickness and demean-age, accounting for group, differ from zero? (this is not the interaction either. this should be the relation between thickness and age across 2 groups.) 3. does the average thickness differe between the two groups? (this should be the group main effect) 4, does the thickness-age correlation differ between 2 groups (*I think this should be what I need*).
If so, I further checked whether the fourth one could survive the corrected threshold using monte carlo stimulation (threshold 1.3, p<0.05). two clusters survived.
But one interesting thing is that the results from above 3 and 4 produced the same results except one thing. 3 gave the hightened thickness of several regions in patients groups compared with controls. In contrast, 4 found the same region but the opposite pattern (decresed thickness in patients compared with controls).
When I add my age as a nuisance covariate, I get the same results 3.
I don't know which one should report. Do my results make sense?
My last question is that I could not find the data file which includes the thickness values from each subject resulting from above analysis because I want to do some correlation analysis with other clinical variables.
I hope that it is clear to you all.
Thank you for your help in advance.
Best, Lickey