I have no problem with 40 percent. It may very well be acceptable. However as I said the sample is biased. I never said anything about sample size. The bias stems from your own selection criteria. You say that was your intention. That's fine. But I'm pointing out how your methods bias the results. The 40 percent should be put in context and in order to do that you need the full range.
Moreover I think I asked you for a definition of a "hit" previously. Your post is silent about that. I understand that there is no perfect definition but if I knew how it was defined at least we would be on the same page.
Thus as I said you did it "to an extent" but imo it's not completely satisfying.
On a side note you can think sample size is trope if you want but it's an important element in all statistics.
Edited 2 time(s). Last edit at 03/23/2019 03:47AM by PHDram.