I deal solely in testable empirical predictions. My assertion is that a methodology in which one does not consider statements which do not generate testable empirical predictions is more efficient, other things being equal, than one which does, at generating accurate testable empirical predictions. And that’s a testable empirical prediction.
I deal solely in testable empirical predictions. My assertion is that a methodology in which one does not consider statements which do not generate testable empirical predictions is more efficient, other things being equal, than one which does, at generating accurate testable empirical predictions. And that’s a testable empirical prediction.