Separating Rewards Into Meaningless Categories Increases Motivation
There’s an ancient riddle that goes something along the lines of, “What would you do for a Klondike Bar?” While psychologists are still searching for a complete answer, a new study (pdf) by USC’s Scott Wiltermuth and Harvard’s Francesca Gino provides a partial answer: Not as much as you would do for half a Klondike bar and half a chocolate covered ice cream square made by Klondike.
Wiltermuth and Gino were interested in how the way we categorize rewards influences our motivation, and so they designed a series of six experiments that followed the same general pattern. Participants performed a task where their performance was based on motivation — for example, the amount of time spent transcribing a text — and they were rewarded for their performance. A weak effort would earn one reward and a stronger effort would merit two rewards. The rewards were a “mix of stationary and food items” that were plainly visible in two buckets.
The experimental manipulation was that people in the first group were allowed to pick any reward from any of the two identical buckets, while people in the second group had to choose their first reward from bucket #1 and could only gain access to bucket #2 if they earned the second reward. Participants could see the rewards before the experiment began and were aware the buckets held the same items. Nevertheless, the fear of losing out on whatever was in bucket #2 led to the latter group to demonstrate increased motivation.
Across six experiments, people were more motivated to obtain one reward from one category and another reward from another category than they were to obtain two rewards from a pool that included all items from either reward category. As a result, they worked longer when potential rewards for their work were separated into meaningless categories. This categorization effect persisted regardless of whether the rewards were presented using a gain or loss frame. Using both moderation and mediation analyses, we found that categorizing rewards had these positive effects on motivation by increasing the degree to which people felt they would “miss out” if they did not obtain the second reward.
The study has a lot of implications for nudging behavior — for example, motivation might increase if employees who work a certain amount get a bonus that includes cash and stock — but the thing that jumps out to me is how poorly designed the school day is for taking advantage of these motivational tendencies. Think about the types of rewards students can obtain by doing well in school. There’s knowledge itself, a grade that signals proof of knowledge, praise from parents and teachers, and increased educational and professional opportunities. But though students may have a few different types of rewards, they all come from the same actions, and they are the same for every single class. If a student does well in math class, there’s not a different type of reward that can be obtained by doing well in biology.
Now imagine a educational model where students learn six different subjects in six different ways and acquire six different types of knowledge. Perhaps there is an apprenticeship in a chemistry lab where the reward for learning is being able to use complex machinery. Perhaps there in an engineering project where the reward is playing in the treehouse you helped build. Perhaps history is learned through teaching lessons to younger children. Here the reward for learning is passing on knowledge to others. The point is that in a school system where every subject essentially confers the same rewards and is learned the same way it’s impossible for students to reap the full benefits of being motivated by different categories of rewards. Until we break out of the institutionalized school model we’ve been using for the last 100 years we’ll never be able to utilize our full arsenal of motivational tools.
Wiltermuth, S., & Gino, F. (2012). “I’ll Have One of Each”: How Separating Rewards Into (Meaningless) Categories Increases Motivation. Journal of Personality and Social Psychology DOI: 10.1037/a0030835