From C. J. Evans, University of Wales
There is one application that several readers might already be familiar with.
Suppose you have some text that is being copied manually from one medium to
another (handwriting to typescript, typescript to publisher’s proofs). You read
through the copy carefully and note any mistakes. A colleague reads through an
unmarked copy independently, and also notes any mistakes.
Some of these mistakes will be found by both people, while some will be found
by one and not the other. From these results you can estimate how many mistakes
were probably missed by both people.
Swansea
