open-discussion > p-FDR seed level correction
Showing 1-2 of 2 posts
Feb 13, 2017 02:02 PM | Joyce Vromen
p-FDR seed level correction
Dear Sir/Madam,
I have conducted a ROI-ROI analysis. For the second-level results I use the option 'threshold ROI-to-ROI connections (by intensity)' and then use 'p-FDR seed-level correction' at 0.05 level.
For the connectivity between ROI A & ROI B, the t-value = 3.93, p uncorrected = 0.0004, the p-FDR = 0.0272
For the connectivity between ROI C & ROI D, the t-value = 3.93, p uncorrected = 0.0004, the p-FDR = 0.0542
How come the p-FDR is different for these two connections whilst the t-value is identical? I understood that the p-FDR seed-level correction corrects for the nr of seed ROIs in the analysis and as such that a given t-value should translate into a specific p-FDR within a single analysis. Are there any other variables that go into computing this p-FDR that I am unaware of?
Thank you for your help.
Best wishes,
Joyce Vromen
I have conducted a ROI-ROI analysis. For the second-level results I use the option 'threshold ROI-to-ROI connections (by intensity)' and then use 'p-FDR seed-level correction' at 0.05 level.
For the connectivity between ROI A & ROI B, the t-value = 3.93, p uncorrected = 0.0004, the p-FDR = 0.0272
For the connectivity between ROI C & ROI D, the t-value = 3.93, p uncorrected = 0.0004, the p-FDR = 0.0542
How come the p-FDR is different for these two connections whilst the t-value is identical? I understood that the p-FDR seed-level correction corrects for the nr of seed ROIs in the analysis and as such that a given t-value should translate into a specific p-FDR within a single analysis. Are there any other variables that go into computing this p-FDR that I am unaware of?
Thank you for your help.
Best wishes,
Joyce Vromen
Feb 14, 2017 11:02 AM | David Kennedy
RE: p-FDR seed level correction
Hi. Can you tell us what software you are using? Is it
CONN? If so, their user group
at http://www.nitrc.org/forum/forum.php?forum_id=1144 is quite
active and helpful.