Correcting for multiple correlations

#1
In my paper I have computed several correlations. I am trying to figure out the most appropriate correction for my p value. Bonferroni seems to be too conservative as I do not want to mask any interesting results/make type 2 errors because my study is explorative.

Does anyone have any suggestions? Can I use a standard 0.01 level?

Thanks!
 

hlsmith

Less is more. Stay pure. Stay poor.
#2
How many correlations did you make? I usually just use Bonferroni, it is straight forward and errors on the side of caution in our period of poor repeatability.

If you did say 6 pairwise comparisons using alpha 0.01 probably wouldn't be a big deal, but if you did a bunch of comparisons it would likely be a little misleading. Also, you should be reporting confidence intervals in lieu of pvalues which aren't as informative.