[rwmj]

Could there be a kind of Taylorism going on here where the effect just happens because it's such a dramatic change to normal, and the effect will go away if the study continued for longer?