Dear FreeSurfers,
I’m comparing two groups (patients (N=45) and controls (N=93)) on cortical thickness using 34 parcelated areas in each hemisphere (i.e. 68 areas in total). When correcting for multiple comparisons using Bonferroni, only one area remains significant. Bonferroni is quite conservative (alpha needs to be lower than 0.05/68= 0.000735). Could you recommend any other, more lenient correction statistics for my data?
In a search through the literature, I encountered adjusted Bonferroni or Benjamini-Hochberg as alternatives. Do you think these would be useful statistics, or have I missed other, more appropriate ones?
Thanks a lot,
Anna
You can use FDR correction.
On 08/14/2013 08:16 AM, Anita van Loenhoud wrote:
Dear FreeSurfers,
I’m comparing two groups (patients (N=45) and controls (N=93)) on cortical thickness using 34 parcelated areas in each hemisphere (i.e. 68 areas in total). When correcting for multiple comparisons using Bonferroni, only one area remains significant. Bonferroni is quite conservative (alpha needs to be lower than 0.05/68= 0.000735). Could you recommend any other, more lenient correction statistics for my data?
In a search through the literature, I encountered adjusted Bonferroni or Benjamini-Hochberg as alternatives. Do you think these would be useful statistics, or have I missed other, more appropriate ones?
Thanks a lot,
Anna
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
Hi Anna, you can use FDR (ie, Ben-Hoc). You can also use gaussian random fields (GRF) to do voxel-wise or cluster-wise stats. In our implementation, we actually use a monte carlo simulation. Run mri_glmfit and spec the label you want to constrain your results to (--label), then run mri_glmfit-sim and specify that you want to do the simulation. doug
On 8/14/13 8:16 AM, Anita van Loenhoud wrote:
Dear FreeSurfers,
I’m comparing two groups (patients (N=45) and controls (N=93)) on cortical thickness using 34 parcelated areas in each hemisphere (i.e. 68 areas in total). When correcting for multiple comparisons using Bonferroni, only one area remains significant. Bonferroni is quite conservative (alpha needs to be lower than 0.05/68= 0.000735). Could you recommend any other, more lenient correction statistics for my data?
In a search through the literature, I encountered adjusted Bonferroni or Benjamini-Hochberg as alternatives. Do you think these would be useful statistics, or have I missed other, more appropriate ones?
Thanks a lot,
Anna
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Anita van Loenhoud -
Chris Watson -
Douglas Greve