where does he say that specifically? He refers to statisticians helping to do further analysis on the data, but I don't see where they showed that the main claim (that the test is predictive of success in the course) is refuted by statisticians.