Hi,
You should use the *corrp* files to report your results as these give you p-values that are corrected for multiple-comparisons. They contain 1-p values, so any values higher than 0.95 are the statistical significant ones (that is, where the corrected p-values are less than 0.05). I would not worry about the tstat files at all.
All the best,
Mark
On 8 Apr 2014, at 14:47, Manuel Blesa <[log in to unmask]> wrote:
> Hi,
>
> Without apply the demean otion the vox_p_tstat file looks nice, is correct? I don't know which file should I use to see my results after randomise. Is correct to see the results using the _tstat file (if I put a range from 1.6 to 6) or I should use _vox_p_tstat (0.95-1), what's the difference between this two files? I'm confused with the results obtained about randomise and I don't know how to see it to make a correct interpretation. Thanks in advance
>
> Best regards,
>
> Manuel Blesa
|