If they assume that the illness affects the population entirely randomly, then comparing the people's performance after the illness to the average makes sense. But if the illness is more likely to affect people who were below average to begin with, then they're just testing people who were already below average.
If they assume that the illness affects the population entirely randomly, then comparing the people's performance after the illness to the average makes sense. But if the illness is more likely to affect people who were below average to begin with, then they're just testing people who were already below average.
(post is archived)