Not hearing something you should be able to is another demonstrated form of confirmation bias, and should be accounted for in good test design. You'll see this in the good published studies. It's not so easy with self-administered tests though. I think quick switching being more sensitive (among...