4 out of 80 labelled uncontaminated are actually contaminated.
Ah!! Is that easy .. It appeared as a hard nut......
Can I get a more detailed explanation ?
Oh it's not that bad!!
I picked #'s, in this case 100 samples.
20 of these are actually contaminated, while 80 are uncontaminated.
If they correctly labelled 80% of the contaminated samples, then that means they labelled 16 out of the 20 correctly, but there are still 4 contaminated ones sitting out there. Now they labelled 5% of the uncomtaminated samples wrong as well, so they marked 4 as being contaminated when they are indeed not. So in the end there are 80 marked uncontaminated, but from the beginning (when they marked only 16 out of the 20 contaminated) there are four of them that are contaminated.