>>14248709thank you, I am glad to know a smart person is out there somewhere
I can't tell you how many projects I have worked on were the data does something like 23/77 or 19/81 and the team just assumes the data is wrong and uses 20/80. It is not that they are just rounding for convenience which I could understand, it is that they treat 20/80 as some holy law and toss out the parts of data that don't match because it must be wrong. I get frustrated give all the careful work my team does to get that data, and that it hurts our final product by a few points we could have gotten if not for being blinded by 20/80. And don't get me started on other stuff like p=0.05.