This would be the second-best answer, but it suffers from a 
slight construct validity problem.  This is because subjects
were never really rewarded in this experiment.  There is
another suggested answer that is slightly more accurate.