Beyond Empiricism. On criteria for Educational Research