“…For example, if the cut o criteria was initially set to give a false positive rate of 10% (90% probability level, see Table 1) then repeating the test with the same criteria, the patient being required to be beyond the 90% probability level on both tests, would reduce the false positive rate to 1% (the square of its value). It is for this reason that several researchers have advocated duplicate measures (Hoskins et al, 1988;Schultzer, 1994) when trying to distinguish between noise and progressive loss. The Normal Tension Glaucoma Study (Schultzer, 1994), which was designed to detect the smallest degree of progression, found that a single stage cut o value produced unrealistically large numbers of progressors.…”