true ... but often (as in the case of this article) the results are presented as if they are a surprise... which implies a null hypothesis anyway I agree with

true ... but often (as in the case of this article) the results are presented as if they are a surprise... which implies a null hypothesis anyway I agree with you, one doesn't need a null hypothesis to measure things on the other hand to advance theories, in other words to test and reject (or fail to reject) hypotheses, one does need ... hypotheses[1] [1]: http://en.wikipedia.org/wiki/Karl_Popper http://en.wikipedia.org/wiki/Karl_Popper