Dear CONN users,
I want to analyze functional connectivity in the DMN with rs-fMRI data of patients and controls. My main hypothesis is focussing on the patient group. I did a pre post measurement.
My patient group was split into 2 subgroups, one that did a control intervention and the other did a physical activity intervention. Now I want to analyze if the intervention had an effect in the patient group that did the PA intervention compared to the control intervention patient group. From my understanding this is a Mixed-Design ANOVA, because we compare different groups (control intv, intv) and different timepoints (pre, post).
Is this possible to analyze in CONN?
I added a screenshot of my analysis!
Sincerely
A psychology student :)
Dear psychology student
Yes, that is a perfectly correct model and contrast to test the interaction in your mixed design model, evaluating whether the effect of intervention (i.e. the difference in connectivity between pre- and post- intervention scans) depends on the type of intervention (i.e. is itself different when comparing subjects that did a PA-intervention vs. subjects that did a control-intervention).
(as a side note: it is not necessary to explicitly enter a [1 -1; -1 1] contrast -evaluting the two directions of the same effect- in these cases, you could as well just enter the simpler [-1 1] contrasts instead and, since CONN default behavior will be to use two-tailed statistics, the results would be identical)
Best
Alfonso
Originally posted by psyolstudent:
Dear CONN users,
I want to analyze functional connectivity in the DMN with rs-fMRI data of patients and controls. My main hypothesis is focussing on the patient group. I did a pre post measurement.
My patient group was split into 2 subgroups, one that did a control intervention and the other did a physical activity intervention. Now I want to analyze if the intervention had an effect in the patient group that did the PA intervention compared to the control intervention patient group. From my understanding this is a Mixed-Design ANOVA, because we compare different groups (control intv, intv) and different timepoints (pre, post).
Is this possible to analyze in CONN?
I added a screenshot of my analysis!
Sincerely
A psychology student :)
thank you Alfonso! :)
Additionally I want to correlate the change in functional connectivity values against a behavioral variable (Increase in step count).
I've read many replies from you and you suggest extracting the values from the result window and use the calculator in conn? So change in FC would be my dependent variable (Y) and step count my indipendent variable (X)
Best wishes and happy new year from Germany!