Gang Wrote:
-------------------------------------------------------
> OK, I didn't notice the following sentence of
> yours:
>
> > Is there a way to get the one-sided upper tail
> p-value?
>
> Then do this:
>
> 3dcalc -a stats.Imagination001_REML+orig'[4]'
> -expr
> 'ispositive(-a)*(1-fitt_t2p(a,20)/2)+ispositive(a)
> *fitt_t2p(a,20)/2' -datum float -prefix
> brick4p_one_sided
Great! Same as my finding, except that the first condition should be 1-ispositive(a) or zeroes will be eliminated from the calculation.
And this "works"! I will now validate it by pulling out the tstats and comparing in R.
Thanks!