ATTENTION: This email originated from a sender outside of MCW. Use caution when clicking on links or opening attachments.
________________________________
Hi, I see in the FslWiki that there is a model for "Two group difference adjusted for covariate" (option1) and "two groups with continuous covariate interaction" (option2). I have a couple of questions regarding these two models.
1. My understanding is that the first one is used to test if FA (my dependent variable) is different between the two groups controlling for some covariate (i.e. age). The second one is used to test if the differences in FA between group1 and group2 vary with
age. Is this correct?
2. The documentation says "if the interaction is significant, only the inferences of the interaction are of interest". Does this mean that if the second model gives significant voxels, I cannot test if there's differences between groups? So, I will never be
able to say "FA is lower in patients than controls" because that difference is affected/depends on age?
