search

dmcartor / MDMR

Multivariate distance matrix regression in R
5 stars 1 forks source link

If the P-value is too small #2

Open liu-psy opened 5 years ago

liu-psy commented 5 years ago

In mdmr(), when I use permutation(perm.p = T) to test my data, if the one predictor's p-value are too small , it will be 0, which I can't correct it in next stage. How can I do to avoid this situation.

Thanks advance!

jrasero commented 4 years ago

Hi @1excel ,

maybe it's too late since you asked this, but if your p-value is 0 when using perm.p = T, it means that the p-value is smaller than 1/nperm. You may need to increase the number of permutations by changing the parameter nperm, which by default it is 500 (telling you that the p-value is < 0.002), in order to get a different from zero p-value.