I am seeking to apply a correction to my exploratory analyses to account for familywise error. I am conducting moderated MRAs with 7 different moderators, 1 IV, and 1 DV. I am interested in interpreting the interaction effects between each of the moderators and the IV on this particular DV.
At an alpha criterion of .05, the Bonferroni adjustment provides a criterion of .05/7 = .0071. At this criterion, one of my interaction effects was barely significant at p = .007. The Holm-Bonferroni adjusted alpha for this same effect also was significant at .049.
I probed this interaction effect using simple slopes, which revealed p = .008 at the low level of this particular moderator. However, I am wondering what alpha criterion I should be using to interpret this effect? Would I interpret this effect at the alpha criterion of .05, or would I also use a Bonferroni or Holm-Bonferroni corrected alpha criterion to interpret the significance level of this simple slope?