 |
 |
Dear Feng, the values FDRp and FDRc and the bottom left of the results table correspond to topological FDR, while I assume that you are looking at results using voxel-wise FDR (after changing SPM defaults setting). Best regards, Guillaume. On 16/04/15 11:06, Feng Liu wrote: > Dear SPM experts, > > I have a question about SPM results table. What are the meaning of FWEp, FDRp, FWEc and FDRc at the left bottom of the table. Given an uncorrected p value at voxel level, the FWEc and FDRc are the cluster size threshold of FWE p<0.05 and FDR p<0.05 cluster level correction, respectively. Is this right? > > The FWEp is the T/F (T test or F test) threshold of FWE p<0.05 peak level correction, right? I performed FWE p<0.05 correction and the corresponding height threshold T = 4.985285, which is equal to FWEp (Attached file 1). I also performed FDR p<0.05 correction and the corresponding height threshold T = 2.385354 (Attached file 2). However, the FDRp is 4.032, which makes me very confused. Moreover, the FWEp in two attached files are equal (4.985285), but the FDRp are different, why? > > Thanks in advance. > > Best, > Feng > -- Guillaume Flandin, PhD Wellcome Trust Centre for Neuroimaging University College London 12 Queen Square London WC1N 3BG