> It seems what I am looking for can best be achieved by
>
> abs(x/sx) < tinv(0.975,df)
That looks right. You can compute the p-value using tcdf, or compute a
cutoff for a significance test using tinv this way.
-- Tom![]()
>
> abs(x/sx) < tinv(0.975,df)
That looks right. You can compute the p-value using tcdf, or compute a
cutoff for a significance test using tinv this way.
-- Tom