Okay Kiwi Spergs, some of you skipped statistics 101.
The null hypothesis is the inverse of whatever your hypothesis is. If your hypothesis is "monkeys like bananas" the null hypothesis is "we will not find that monkeys like bananas".
Type 1 errors, in short are false positive results of a test. In more technical terms, the null hypothesis is rejected and it is erroneously affirmed that there is a statistical difference.
Type 2 errors are false negatives. The null hypothesis is erroneously affirmed.
Drug testing optimizes for
type 2 errors, not type 1 errors. Meaning, they are highly optimized to only incorrectly. I found
this study that says when it comes to cannabis follicle testing, they expect
0 (with 95% confidence) instances of false positive (meaning the person was erroneously found to have drugs in their system when they did not), and many more false negatives.
I would expect the balldowashers to come out in droves to say "'nooo follicle testing has an x% chance of being wrong" while completely misunderstanding (or lying) about the fact that those are false negatives, not false positives.