This relationship between reliability and sample size indicates that there are, to use a phrase from economics, diminishing returns to increasing the size of a sample. It seldom pays to take samples that are massively large.
I am not sure whether it was Freund or Perles who came up with this. I guess one of them probably had taken a course in economics before; or he had done research in economics, which is vastly probable - at least there is a sub-discipline in economics called Bayesian.
No comments:
Post a Comment