Intrinsic Motivation, Extrinsic Motivation and the Process of Learning

Bing Nursery School


Professor Mark Lepper, Bing Distinguished Lecture Series

By Christine VanDeVelde

Some years ago, after a lecture, Professor Mark Lepper was approached by a couple who told him about a system of rewards they had set up for their son, which had produced much improved behavior at the dinner table. "He sits up straight and eats his peas and the Brussels sprouts and he is really very well behaved," they reported. Until, that is, the first time the family dined at a nice restaurant. The child looked around, picked up a crystal glass from the table and asked, "How many points not to drop this?" A fine example, says Dr. Lepper, of the detrimental effects of over-reliance on rewards to shape children's behavior.

On May 29, Dr. Mark R. Lepper, Chairman of Stanford's Psychology Department, presented "Intrinsic Motivation, Extrinsic Motivation, and the Process of Learning," an overview of the research into the role rewards play in children's learning at the annual Bing Nursery School Distinguished Lecture.

At the time Lepper began his research, behavior modification programs were in their heyday. If kids were misbehaving, not studying enough, not learning, then the answer was, Lepper says, "good old-fashioned capitalism." The children should be paid -- with rewards, points, chips -- for good behavior. Points could be redeemed for extra recess time or candy or even to buy your way out of tests. Conversely, points would be taken away for bad behavior. (That's 50 points from Gryffindor!)

Such programs produced an instant effect. The rewards were very attractive and the children shaped up instantly. And as long as the program continued, so did the students' behavior. That, says Lepper, was the wonderful positive effect and the reason for popularity of these programs.

But, Lepper asked, what happens when these children leave school? What happens when they go to a public school or a school that doesn't have this kind of program? In fact, when such programs were discontinued, the children sometimes got much worse. They would work to get rewards, but if there was nothing in it for them, then there was no reason at all to pay attention or behave or study.

While looking more closely at a number of these programs -- some of which had gone so overboard they were simply out and out bribery -- Lepper observed that while the rewards produced instant compliance, they simultaneously seemed to undermine real motivation. There were hints in the literature of these effects, but no one had really studied them.

So, Lepper designed a series of studies to explore these effects. (Such a study, by the way, could only be done at a research school like Bing.) Over a three-week period, during the first hour of each class, the teachers put out on one particular table in front of the one-way observation mirrors a new activity--magic markers and drawing paper. Each day, when the children arrived during free play time, this was one of the many choices they had.

From behind the one-way mirror, the researchers could measure how much time during these free play periods each of the children chose to spend with this activity as opposed to others. The participants selected for the study were only those children who showed a high level of interest in the activity--in other words, children who were already intrinsically motivated.

Then, each of those children was taken into one of the game rooms, where they were asked to engage in the same activity under three different conditions. Under one condition, a reward was offered; the child saw in advance the "Good Player" Award with its line for their name and agreed to draw with the magic markers in order to get it. Under another condition, when the child finished their drawing, they were unexpectedly given a reward. In the third group, children neither expected nor received any tangible reward, but did receive the same feedback on their work as the other children.

Two weeks later, the teachers again put out the magic markers in the classroom. From behind the one-way mirrors, the experimenters observed how much time the children chose to spend with the activity, when there was no longer any tangible reward available.

What happened? The children who had contracted to receive the "Good Player" award showed significantly less interest -- in fact, half as much -- as they had before the study. So, contracting for a reward to do this initially interesting and attractive activity subsequently had a negative effect on their interest. The misuse of rewards or the use of superfluous awards undermined intrinsic interests, turning an attractive activity into something the child would only want to do if there was a payoff.

Later studies varied conditions and rewards, but, the same basic effect was always noted -- children expecting the reward during the experimental session showed less subsequent interest in the classroom and less interest than they had initially. In a related group of studies, the same effect was found when children had to complete activities under tight time deadlines. And in yet another related study, children's art teachers were asked to rate the creativity, quality, and interest value of the paintings done during the experiments and found that children who were expecting the reward drew more pictures but of lower average quality.

So, how is it that children learn the principle that when somebody offers you a bribe for doing something, then what you're being offered the bribe for is probably something that you're not going to find very interesting? Lepper suggested looking at the classic dinner table debate, when Mom or Dad says you can't have your dessert until you've finished your vegetables. Does the debate itself -- the "means-end" condition -- tell the children how they should feel about dessert as opposed to vegetables?

Lepper and his team set up an experiment in which this dinner table scene was described to forty four-year-old children at Bing, substituting the fanciful food names: gumblatts and snogworts. "We actually have samples of these two foods," they told the children. "They're under covered dishes so you can't see or smell them. This one's the gumblatts (the one you couldn't have right away at the dinner table -- the end) and this one's the snogworts (the one you had to eat to get the other one -- the means). Which one would you like to try?"

Ninety percent of the children immediately went for the gumblatts. Even the ten percent who didn't all still showed that they understood the social script -- remarking that the snogworts probably had more vitamins and minerals or the gumblatts had "more refined sugars" (a four-year-old participant whom Dr. Lepper dearly wished he had followed up on to see how he fared in later life). The upshot is that when children are told they have to do one thing in order to do the other, the thing presented as the reward becomes more attractive and the thing done to get the reward is less attractive.

All of this, however, does not mean that extrinsic rewards should never be used. Rewards, in fact, are neither good nor bad. There are good and bad uses of them. So, how can rewards be used most effectively, without undermining interest or in order to achieve other positive ends? It comes down to looking at rewards and intrinsic motivation from the standpoint of four questions, says Lepper.

First, is intrinsic motivation relevant? For example, you might have a rule that your child take out the garbage, but you don't expect that he will learn to enjoy it. You're just happy if he does it on time and without being prompted. So, if offering some kind of reward is helpful in such a circumstance, there is no problem.

Second, is the reward necessary or superfluous? Is the reward needed to get the children to engage in the activity? In the original studies, activities were deliberately selected for which the reward was superfluous--the child was motivated to do the activity anyway without reward. But when a child will not engage in an activity without some reward, they may have a positive effect by getting children started in the task. Lots of tasks at first can be awful and dull and boring until you acquire enough competence to do them well, like the early stages of reading. If you start by doing things that encourage the child to engage in the activity enough to come to appreciate that it's fun -- even if those things are a little heavy-handed -- that may be a legitimate use for rewards.

Third, does increased engagement in the task help build new skills? In the case of reading, if you can get kids to engage in activities that are relevant to eventually developing this skill, then in the end they will have acquired a skill that has intrinsic value that they didn't have before.

Finally, if you have to use rewards in cases where you want to encourage children to do something, will the child perceive the reward as a bribe or a bonus? Ideally, there are times when rewards focus the child on a feeling of competence and accomplishment and can be used when we want to convey pride in a child's accomplishments. That's very different from using rewards for social control or as a technique of discipline.

As Lepper noted, the British philosopher John Locke first observed in 1693 in Some Thoughts on Education that in teaching a child, care must be taken that learning never be made a business to him. "I've always had the fancy that learning might be made a play and a recreation to children, they might be brought to a desire to be taught if only learning were proposed to them as a thing of delight and recreation," Locke wrote. Teachers -- and parents -- often unwittingly turn play into work, the source of a further unintended effect on intrinsic motivation. The longer children are in school, the less they seemed to be intrinsically motivated. Certainly, trappings such as grades and test scores become more important as children progress in school, but overall, such extrinsic motivation or rewards, stay fairly level. So, the final direction Lepper's research took was a look into how to turn work into play and led to what he calls "The Five C's."

The first "C" for turning work into play is challenge. There is a lot of evidence that children -- as well as the the rest of us -- will seek out challenges, that if you give children tasks of different levels of difficulty, they'll look for one of intermediate difficulty -- the one where they're not certain they're going to succeed, but it's not impossible. They think they can improve and learn and become better.

It's fun.

For the next "C," remember children search for competence, evidence that they've accomplished something at a high level, or that they've improved. And they like to feel that they are personally responsible for their success, that it wasn't just luck or the ease of the task. Only when the task is challenging do they begin to feel competence when they succeed, when they feel like effort, skill and ability entered into the success.

The third "C" is that people of all ages like to be in control. They like to feel like they're in charge, that they're determining their own fates. This is a particularly American or Western European concept.

The fourth "C" is curiosity. We often seek out things because we're curious, because they're mysterious and complex, things we sort of understand but not quite. The incongruity makes us want to learn more. Good teachers are adept at bringing out this sense of wonderment.

And finally, the fifth "C" is context, which refers to the fact that we often get great pleasure from engrossing ourselves in imaginary environments -- listening to stories, reading books, going to movies, watching TV, playing video games. It's not clear in the literature, Lepper says, precisely what the rewards are of identifying with characters, but it's clearly a very powerful effect.

The ultimate question is: do children learn better or differently when they're intrinsically motivated? To find out, Lepper set up a study where a computer-based learning activity was presented as a game. The children who were participating were given a series of choices about aspects of the activity and allowed to personalize the game.

Three effects were observed. First, children were more interested and learned more when the activity was made into a game. They also learned more when they had a choice over trivial aspects of the game -- such as choosing their game piece or naming their game character. Finally, it was found that in increasing the intrinsic aspect of the task, better learning resulted -- the children got the answers more quickly. Extrinsic motivation interfered with learning. Further, it was found that intrinsically motivated children showed higher levels of perceived ability, causing them to desire more difficult tasks, basically generalizing the process of learning to other contexts.

In closing, Lepper pointed out that if you could create a child who was perfectly intrinsically motivated and absolutely uninterested in every extrinsic motivation, that person probably would not succeed very well in our society. In the final analysis, like many aspects of child-rearing, intrinsic motivation and the use of rewards are a balancing act.

Looking back over the span of all this research, Lepper feels he ultimately discovered what Mark Twain had found a hundred years earlier in The Adventures of Tom Sawyer, where Tom conned the other kids into whitewashing Aunt Polly's fence by telling them it was such a special privilege to be able to do so. Twain wrote, "For Tom had discovered a fundamental law of human nature, that work is what a body is obliged to do, and that play is what a body is not obliged to do. And this should help explain why constructing artificial flowers or walking on a treadmill is work, whereas bowling ten pins or climbing Mont Blanc is merely amusement." It's kind of like the difference between gumblatts and snogworts.