Testing vs. believing hypotheses: magical ideation in the judgement of contingencies