The danger of over-explanation in psychological research: A reply to Griffiths