External Email - Use Caution
Dear Freesurfers,
I am conducting a study on the link of cerebral structural measurements and psychopathological dimensions. At first, I investigated the link of structural markers and the respective dimension while controlling for the patient's age and the other dimensions, that are investigated in my study. As I will maybe focus on only one of the dimensions, I ran exactly the same analysis for this dimension again, while only controlling for age. Suprisingly, this analysis does not show any significant correlations, while the analysis which additionally controlled for the other dimensions showed several clusters of significance in this dimension. I assumed the analysis with less variables of control would rather find more clusters, as less variables of control are used. Am I mistaken? Have any of you encountered this problem before or can explain, why less variables of control lead to fewer clusters with significant correlations? Any help is appreciated!
Best regards
Kai Ohmstedt