problem with the scientific method