Think of another situation, like the music students and the extra tuition example, where you would believe that the benefits were "for real" regardless of results from a statistical analysis telling you that the favorable results could be easily explained by chance. POLI967