help
help > RE: How to use "age" as covariates?
May 29, 2014 08:05 AM | Alfonso Nieto-Castanon - Boston University
RE: How to use "age" as covariates?
Hi Yifei,
The connection-level p-FDR (seed-level correction) performs only multiple-comparison correction across multiple target ROIs for a single seed ROI (or for each seed-ROI separately, if multiple seeds are selected), so it only allows you to make appropriately corrected inferences when you have a single seed-ROI of interest (or when you have individual hypotheses about each one of several ROIs).
The seed-level threshold correct across multiple seed ROIs (so it only appears when selecting more than one source/seed ROI). Similarly the network-level threshold applies only to full multiple-ROI connectivity graphs so again you need to select more than one source/seed ROI for this option to become available. When you select a single source/seed ROI and enable permutation tests you still get the associated statistics for the chosen seed listed on the stats table (e.g. size or intensity of the connections between this ROI and all other ROIs). You just do not need to apply any additional corrections across multiple seeds so the associated seed-level thresholding options do not show up in the define threhsolds menu.
Hope this helps/clarifies
Alfonso
Originally posted by Yifei Zhang:
The connection-level p-FDR (seed-level correction) performs only multiple-comparison correction across multiple target ROIs for a single seed ROI (or for each seed-ROI separately, if multiple seeds are selected), so it only allows you to make appropriately corrected inferences when you have a single seed-ROI of interest (or when you have individual hypotheses about each one of several ROIs).
The seed-level threshold correct across multiple seed ROIs (so it only appears when selecting more than one source/seed ROI). Similarly the network-level threshold applies only to full multiple-ROI connectivity graphs so again you need to select more than one source/seed ROI for this option to become available. When you select a single source/seed ROI and enable permutation tests you still get the associated statistics for the chosen seed listed on the stats table (e.g. size or intensity of the connections between this ROI and all other ROIs). You just do not need to apply any additional corrections across multiple seeds so the associated seed-level thresholding options do not show up in the define threhsolds menu.
Hope this helps/clarifies
Alfonso
Originally posted by Yifei Zhang:
Hi Alfonso,
Thank you so much for the explanation, It clarified a lot! However, I am still a little bit confused on how to make inference when only use the p-FDR(seed-level correction) threshold. And how to enable the permutation test when selecting only one ROI? (there was no seed/NBS-level correction drop menu when I selected only one ROI and click the button of "enable permutation test")
I am using the conn 14.c, SPM8(4010) and Matlab 2012b versions on Mac.
Thank you!
With best regards,
Yifei
Thank you so much for the explanation, It clarified a lot! However, I am still a little bit confused on how to make inference when only use the p-FDR(seed-level correction) threshold. And how to enable the permutation test when selecting only one ROI? (there was no seed/NBS-level correction drop menu when I selected only one ROI and click the button of "enable permutation test")
I am using the conn 14.c, SPM8(4010) and Matlab 2012b versions on Mac.
Thank you!
With best regards,
Yifei