Hi all: I'm performing the group analysis on my own data using command line on FS5.1.0 and i'm a little confused about the Clusterwise Correction for Multiple Comparisions. the command is as follows: mri_glmfit-sim \ --glmdir lh.gender_age.glmdir \ --sim mc-z 10000 4 mc-z.negative \ --sim-sign neg --cwpvalthresh 0.05\ --overwrite In this command, does the vertex-wise threshold (here, '4') mean that the vertex whose p value is lower than 0.0001 can be considered to used for the Correction for Multiple Comparisions ? likewise, does the sign (here,'neg') mean that only negtive vertexs can be considered? Is there a criterion which vertex-wise threshold and sign i should choose? Hope you can give me some suggestions !
Those vertices with p-value lower than 0.0001 are used to form the clusters for the cluster-wise correction. If the signal in your SPM is weak but extended you should use a permissive cluster-forming p-value (eg 0.01). On the other hand, if your signal is strong and very localized you should use a more stringent p-value. The sign parameter simply tells the command if you want to perform a one-tailed test. For instance, if you are sure that group1 has reduced cortical thickness in comparison with group 2 then you can test the null hypothesis that group2=group1 against the one-tailed alternative hypothesis group2>group1.
-Jorge
De: ZhiLiangLong lagosslong1987@163.com Para: freesurfer freesurfer@nmr.mgh.harvard.edu Enviado: Domingo 18 de marzo de 2012 4:46 Asunto: [Freesurfer] question about Cluster-wise Correction for Multiple Comparisions
Hi all: I'm performing the group analysis on my own data using command line on FS5.1.0 and i'm a little confused about the Clusterwise Correction for Multiple Comparisions. the command is as follows: mri_glmfit-sim \ --glmdir lh.gender_age.glmdir \ --sim mc-z 10000 4 mc-z.negative \ --sim-sign neg --cwpvalthresh 0.05\ --overwrite In this command, does the vertex-wise threshold (here, '4') mean that the vertex whose p value is lower than 0.0001 can be considered to used for the Correction for Multiple Comparisions ? likewise, does the sign (here,'neg') mean that only negtive vertexs can be considered? Is there a criterion which vertex-wise threshold and sign i should choose? Hope you can give me some suggestions !
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.
ZhiLiangLong wrote:
Hi all: I'm performing the group analysis on my own data using command line on FS5.1.0 and i'm a little confused about the Clusterwise Correction for Multiple Comparisions. the command is as follows: mri_glmfit-sim \ --glmdir lh.gender_age.glmdir \ --sim mc-z 10000 4 mc-z.negative \ --sim-sign neg --cwpvalthresh 0.05\ --overwrite In this command, does the vertex-wise threshold (here, '4') mean that the vertex whose p value is lower than 0.0001 can be considered to used for the Correction for Multiple Comparisions ?
yes
likewise, does the sign (here,'neg') mean that only negtive vertexs can be considered?
yes
Is there a criterion which vertex-wise threshold and sign i should choose?
If you have an apriori hypothesis as to the sign that your results will be (eg, male>female), then you can use that sign. If not then use abs (absolute). I would not choose a vertex-wise threshold less than p=.05.
Hope you can give me some suggestions !
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Douglas N Greve -
jorge luis -
ZhiLiangLong