Hi Andrew,

I ran a t-test with FDR/100000 permutations and a p-value= 0.05.
I got an edge which is significantly different between my two groups, and when I click on it, it gives me information about the name of the nodes and then the test stat = 4.12. 

I know that probably this is a question simple to answer, but does the test stat indicate the Wald test value or the t-value? I mean can I use that value to obtain my exact p-value, or is my exact p-value the alpha that I set at the beginning? (in my case has my connection been found significant at a p-value=0.05?). It is just for my own clarification and also to be sure on how to report my results.

Thanks for taking the time to answer this question.

Best regards,
Fran

Hi Fran,

The test stat indicates the t-value. The p-value is exact in the sense that it is derived using permutation.

The p-value that you set at the beginning with alpha is just the threshold of significance. If you set alpha=0.05, then any results with p-values below 0.05 will be considered significant.

Yes you can assume that you edge has been found significant at p<0.05.

You may want to check with 200000 permutations to ensure that the result remains, although 100000 should probably be enough.

Andrew


Originally posted by francesca sibilia: