question about BH correction

[toma...@gmail.com]

Hello,

I have just discovered compareGroups and I am using it for a paper I am currently writing. Also, I am new to these sorts of analyses, so I apologize in advance if this is a sort of "easy" question.

In my analysis, I am comparing the values of 3 categorical variables across 4 groups. When I conduct the comparisons and obtain the p-values for the bivariate comparisons between pairs of the 4 groups, I understand that the package automatically corrects those p-values using the BH adjustment. This means that the p-value I observe when I use the createTable command are alreadu adjusted values, correct?
My question is the following, if I look at those adjusted p-values, is it safe for me to say that any value <0.05 is significant at the 0.05 level? Or do I have to manually obtain the BH Critical value for this set of p-values and then identify which are significant based on that BH critical value?

Thank you for your time.

[compareGroups]

First of all, thanks for using compareGroups package. Hope it will be usefull for you.

About your question, comparGroups function has an argument called p.corrected which performs BH correction when displaying pairwise comparisons p-values for categorical and continuous non-normal distributed variables. This argument, "p.corrected", is set to TRUE by default. Therefore, p-values displayed after calling createTable can be compared to nominal significance level (0.05) without having to do any calculation by hand.

Regards,

Isaac Subirana.