Dear all masters :
We have a set of data ( n=5) with interictal and ictal SPECT. In
order to find the possible rCBF increase during ictal status, we
compared 2 set data using pair t test using SPM2. But under the Sun
Matlab 6.0 , the SPM2 seems stopped during pair t test without any
results. Then we used SPM99 to run and we can proceed to statisitc
analysis.
In some papers , the threshold was set to be 0.0001 and some was
set to 0.01. I want to ask is there any test for us to choce the
threshold level ? I had read FDR method, but I can't find FDR option
from SPM2 manual. Should I load .MAT file generated from pair t test
and then run FDR under matlab in order to find the p ?
And due to lack of knowledge here, is there paper or website about
uncorrected p and corrected p defination ?
Thanks very much !
|