The man in the “surprise room” had a card in his hand. It was about the size of an index card, with the words Good Player Award, a big gold star, a bright red ribbon, and space for a name and a school.
Across a child-sized table sat a preschooler from the Bing Nursery School on the Stanford campus, somewhere between three and five years old. On the table were magic markers, the felt-tip pens the children didn’t normally get in class. A few days earlier the markers had been set out in the classroom, and observers behind a one-way mirror had timed who sat down to draw. This child had been one of the keen ones.
“Would you like to win one of these Good Player Awards?” the man asked. The children agreed, the researchers wrote, “typically with considerable enthusiasm.” Each drew for about six minutes, got an award with their name written on it, and watched it go up on an Honor Roll board.
Nothing about this was unusual in 1973. Schools ran on gold stars and grades, and a newer method, the “token economy,” paid children in tokens for good behavior. The logic was the same throughout: reward what you want more of.
One to two weeks later, the markers went back on the classroom table. Nobody mentioned awards. Behind the mirror, observers who didn’t know which child had been in which group started timing again.
The children who had drawn for the award spent 8.59% of their free time with the markers. Children who had drawn without any award spent 16.73%.
Same star, different deal
There was a third group. These children got the very same card and gold star, with their names on the same Honor Roll. The only difference was that nobody had told them about it in advance. Back in the classroom, they spent 18.09% of their time with the markers, no less than the children who got nothing.
So in this experiment the award on its own showed no sign of damage. What seemed to matter was drawing in order to get it.
Promise a prize for something people already enjoy, and they may come to enjoy it less.
Mark Lepper, David Greene and Richard Nisbett built the test on Daryl Bem’s self-perception theory: people work out their own motives much as they work out a stranger’s. See a colleague staying late only in the week bonuses are calculated, and you draw a conclusion. When an outside reason is “salient, unambiguous, and sufficient” to explain the behavior, the theory says, people credit that reason, even in themselves. A prize contracted in advance gives an activity more justification than it needed, which is where the name comes from. The child may conclude that she drew for the star.
Two years earlier, Edward Deci had seen something similar with college students paid to solve puzzles, though, as Lepper’s team pointed out, his result was only marginally significant. Deci and his colleagues later explained such results as rewards experienced as control, the line of thinking that grew into Self-Determination Theory. The two accounts differ on why. They point at the same kind of reward.
A small, careful finding
Lepper’s team warned against stretching it. “There is nothing in the present line of reasoning or the present data,” they wrote, “to suggest that contracting to engage in an activity for an extrinsic reward will always, or even usually, result in a decrement in intrinsic interest.” They had chosen an attractive activity and children who already liked it. Where interest starts very low, or where an activity only becomes enjoyable after some skill, they suggested rewards may be what gets people involved at all.
The argument that followed ran for decades. In 1994 Judy Cameron and David Pierce pooled 96 studies and concluded that, overall, reward does not decrease intrinsic motivation. They found one exception: expected tangible rewards given simply for doing a task, which had what they called a minimal negative effect on time spent once the reward was removed. That describes the Good Player Award. In their own meta-analysis of 128 experiments, Deci, Richard Koestner and Richard Ryan found the same kind of reward undermined interest clearly, and found no such effect from unexpected rewards. The camps disagree about how big the effect is. Their analyses locate it in the same corner.
The leaderboard
Say Ines works in customer support, and in quiet hours she writes help-center articles. Nobody asked her to. She likes explaining things, and the team’s repeat tickets have dropped since she started. Her manager wants the whole team writing, so she plans a leaderboard: points for each article published, a gift card every month for the top three, the ranking shown at the team meeting.
For Ines, that is the Good Player Award. Writing used to be something she chose; now it comes with a reason printed on the screen. The research doesn’t predict that she’ll stop. It suggests her interest may shrink, and in the experiment the drop in interest was measured later, when the markers came back with no award attached.
For teammates who have never written an article, Lepper’s caveat cuts the other way. Points may be what gets them to try a first one.
The manager could treat the two cases differently. Points as a way in for the newcomers. For Ines, no deal at all, only a remark in the team meeting after one of her articles clears a week of repeat tickets, with no rule promising the next one. A surprise only works while it is one. Announce “best article of the month” and it turns back into a contract.
In the Bing classrooms, the paper and markers were back on the same table. The children who had sat down to draw for a gold star now spent about half as much time there as the children who had drawn for nothing.