They did point out that the first round selection is done based on anonymized applications.
It might not be not perfect, but if round 1 is blind to X, and rounds 2 and 3 admit people in similar ratios to round 1, it sounds pretty good to me.
It might not be not perfect, but if round 1 is blind to X, and rounds 2 and 3 admit people in similar ratios to round 1, it sounds pretty good to me.
That sounds to me like the point where blind review of applications would take place.
Ultimately a quality measurement at the end of the process would be the best, however - measure the outputs rather than the inputs.