I hadn't seen that there had been another study specifically set up to show the impact of tester reliability on results to the test. Ah! Glad to see I am not the only one to think there were issues in the original experiment and its interpretations...