13 June 2016
Your LARPing on the internet as not just a stats major, but one who is about to graduate, & you really can’t understand it?
Let me spell it out: 1/Necessary power on any test needs to be calibrated to the underlying rate of what you want to find.
2/In any test there is a trade-off between error-types. A more sensitive test produces more false-positives, and vice-versa.
3/To get the necessary statistical power to find blood with a x770 risk of infection, you need a vastly more sensitive test.
4/A more sensitive test will result in throwing out far more blood because of false positives. And in the end…
5/1% of the population cannot donate enough blood to make up for throwing out many more false positives from 99% of the pop.
6/Unless we strap you all up as blood-bags, Mad-Max style.
7/(& on top of that, there are always new blood-borne diseases, & fags get them first, & it will be a repeat of the 1980s)