in carlsmith and festinger's study of cognitive dissonance people that were paid 20 dollars to tell naive subjects that the experiment was interesting when it was actually boring showed