help
help > group summary r-value from ROIto ROI analysis
May 26, 2017 03:05 PM | Scott H
group summary r-value from ROIto ROI analysis
Hi Alfonso,
I am running an ROI to ROI analysis between two groups. My connectome ring has multiple ROI to ROI pairs that are significant (for example, the connectome ring might look similar to the Fig on pg 28 in the CONN manual). The results tell us there is a significant difference between the two groups (e.g., young > older adults), but we don't know whether these pairs represent greater positive connectivity in young adults relative to older adults, or whether there is less anti-correlation in young relative to older adults. One way to get the answer is to export the connectivity for each ROI pair for each subject and plot them, but that seems cumbersome for so many ROIs. Is there is a more efficient method available (given so many ROIs showing a significant difference), such as getting a group summary r-value for each ROI pair (e.g., ROI1: young adult r = .51, OA r= .17 ; ROI2: young adult r = -.05 , older adult r = -.42) so that one could easily estimate the overall pattern of the connectome ring. Just wondering if this is an option available in CONN that I have not discovered or if you have any ideas for efficiently extracting the sign of the average correlations for each pair for young and older adults?
Thanks,
Scott
I am running an ROI to ROI analysis between two groups. My connectome ring has multiple ROI to ROI pairs that are significant (for example, the connectome ring might look similar to the Fig on pg 28 in the CONN manual). The results tell us there is a significant difference between the two groups (e.g., young > older adults), but we don't know whether these pairs represent greater positive connectivity in young adults relative to older adults, or whether there is less anti-correlation in young relative to older adults. One way to get the answer is to export the connectivity for each ROI pair for each subject and plot them, but that seems cumbersome for so many ROIs. Is there is a more efficient method available (given so many ROIs showing a significant difference), such as getting a group summary r-value for each ROI pair (e.g., ROI1: young adult r = .51, OA r= .17 ; ROI2: young adult r = -.05 , older adult r = -.42) so that one could easily estimate the overall pattern of the connectome ring. Just wondering if this is an option available in CONN that I have not discovered or if you have any ideas for efficiently extracting the sign of the average correlations for each pair for young and older adults?
Thanks,
Scott