help
help > RE: second-level analysis in Conn
Sep 3, 2015 01:09 AM | Alfonso Nieto-Castanon - Boston University
RE: second-level analysis in Conn
Hi Diana,
From the anyEffect_Groups_socialCope.png figure, and for the selected source/target ROIs displayed there, there does not appear to be any significant association between functional connectivity strength and behavioral scores within any of your three subject groups (this is indicated by the last three effects error-bars crossing the y=0 axis). You may test whether that is the case using the between-subjects contrast [0 0 0 1 0 0;0 0 0 0 1 0; 0 0 0 0 0 1] (this tests whether there is any significant association between connectivity and behavioral scores in any of the three groups). If you want to focus on those regions that show significant associations with behavioral scores in any of the three groups (the main effect of the behavioral score factor) use the contrast above to test specifically for these effects. If you want to focus on those regions that show significantly different associations with behavioral scores between your three groups (the interaction between group and behavioral scores) use instead the contrast [0 0 0 1 -1 0;0 0 0 0 1 -1].
Hope this helps
Alfonso
Originally posted by Diana Parvinchi:
From the anyEffect_Groups_socialCope.png figure, and for the selected source/target ROIs displayed there, there does not appear to be any significant association between functional connectivity strength and behavioral scores within any of your three subject groups (this is indicated by the last three effects error-bars crossing the y=0 axis). You may test whether that is the case using the between-subjects contrast [0 0 0 1 0 0;0 0 0 0 1 0; 0 0 0 0 0 1] (this tests whether there is any significant association between connectivity and behavioral scores in any of the three groups). If you want to focus on those regions that show significant associations with behavioral scores in any of the three groups (the main effect of the behavioral score factor) use the contrast above to test specifically for these effects. If you want to focus on those regions that show significantly different associations with behavioral scores between your three groups (the interaction between group and behavioral scores) use instead the contrast [0 0 0 1 -1 0;0 0 0 0 1 -1].
Hope this helps
Alfonso
Originally posted by Diana Parvinchi:
Originally posted by Diana Parvinchi:
> ( Hi Alfonso,
> As you suggested, I selected my groups and then scores in the �Between-subjects contrast�. Namely, I selected age-group1, age-group2, age-group3, age-group1-scores, age-group2-scores, age-group3-scores. then, I selected �any effect of interest and one ROI (Medial Prefrontal Cortex). I have attached a copy of one of the tables which shows all six regressors and connectivity between MPFC and Paracingulate Gyrus left. How would you interpret this graph? Many many thanks for your help!
Best,
Diana.
> )
> As you suggested, I selected my groups and then scores in the �Between-subjects contrast�. Namely, I selected age-group1, age-group2, age-group3, age-group1-scores, age-group2-scores, age-group3-scores. then, I selected �any effect of interest and one ROI (Medial Prefrontal Cortex). I have attached a copy of one of the tables which shows all six regressors and connectivity between MPFC and Paracingulate Gyrus left. How would you interpret this graph? Many many thanks for your help!
Best,
Diana.
> )