Quote Originally Posted by ganulv View Post
You could do a case study of a single nation.
That would be statistically totally insignificant and irrelevant, though. You need statistical significance to find a correlation - otherwise you're merely discussing a random result.

Few serious people will consider a study interesting if it lacks statistical relevance.


He can formulate a hypothesis, of course. He just cannot test it properly without obeying basic statistical principles. Anecdotical evidence can easily be countered with conflicting anecdotical evidence - which means that anecdotical evidence is useless.