I received an interesting suggestion about the terminology I used in chapter 4 of my Unit Testing book, where I wrote about false positives and false negatives.
Together, these two components are a measure of the test’s accuracy — how good the test is at providing signal and eliminating noise.
False positive/negative is an established terminology from the field of statistics, and, because I tried to reuse existing terminology as much as possible, I went with it.
False positive, in particular, means that the test raised an alarm (failed), but that alarm was false (meaning that the system worked properly, contrary to what the test said).
I think you can see why this taxonomy is confusing. Intuitively, the first thing we think of when we hear positive is pass, not failure.
The reader proposed an alternative terminology that could help eliminate this ambiguity. Instead of false positive, I could use false failure, and instead of false negative — false pass.
And even though false pass/failure are not scientifically precise terms, I now like them better than false negative/positive. They are unambiguous and intuitive — much easier to grasp and remember.
I wish I received this comment when the book was still in MEAP. False positives/negatives are one of the core concepts in the book and simplifying understanding of this concept would be helpful.
Well, something to work out in the 2nd edition of the book, I guess.
--Vlad