This blog post is another in a relatively long string of posts describing why I do not feel reward or incentive-based management systems are appropriate models for school-wide behavioral management (previous posts here). Up until now, I have been addressing the issue from an educational perspective with a particular focus on how rewards can turn into punishers over time.
For this post, I will return to my graduate school training and address the issue as a traditional psychologist and explore a theory from the 1960s and 1970s. This theory tries to explain the paradoxical findings on how rewards reduced motivation. I am going to focus on this theory because I often feel like reward-based management systems do, in fact, bribe students into good behavior and teaches children that they should be good for the sake of receiving a reward, and not to be good for goodness’ sake.
Here is a diagram depicting the different ways we can be motivated. On the far right is what we want to be: intrinsically motivated. On the far left is the condition wherein we have zero motivation. In the middle is the spectrum of external incentives that influence our behavior. The further to the left the items are, the less they are likely to support intrinsic motivation. And most relevant to this post, “external regulation” is a synonym to “reinforcement” or “rewards” provided by other people to motivate.
Here is another way to look at intrinsic and extrinsic motivation. It also gives us some terms that help understand what intrinsic motivators might entail.
The hypothesis I alluded to above is called the overjustification effect. It seeks to explain why providing an external reward to an individual for performing a task they are intrinsically motivated to complete results in overall decreased motivation and often less task performance than would happen if the reward were absent.
As stated on Wikipedia (emphasis is mine):
The overjustification effect occurs when expected external incentives such as money or prizes decreases a person’s intrinsic motivation to perform a task. Overjustification is an explanation for the phenomenon known as motivational “crowding out.” The overall effect of offering a reward for a previously unrewarded activity is a shift to extrinsic motivation and the undermining of pre-existing intrinsic motivation. Once rewards are no longer offered, interest in the activity is lost; prior intrinsic motivation does not return, and extrinsic rewards must be continuously offered as motivation to sustain the activity.1
Another operational definition of the overjustification effect that aligns with behaviorist approaches was provided more recently by Ulber, Hamann, and Tomasello (2016) (emphasis mine):
[The overjustification] effect is an implication of self-perception theory (Bem, 1972), proposing that a person’s inference about their own behavior is based on the perception of sufficient external contingencies. If individuals are induced to engage in an activity in order to receive a reward, they often conclude that their actions were primarily motivated by the external incentive rather than by any intrinsic interest in the activity itself (Lepper, 1981; Lepper et al., 1973). Another possibility is that they perceive the external reinforcement as a coercive force, controlling or bribing their behavior (Deci, 1975). As a result, the new extrinsic motivation replaces the initial intrinsic motivation, so that when the extrinsic reward is no longer forthcoming, the motivation for the activity decreases.
A real-world example of the overjustification effect was proposed by Edward Deci in 1971 in his first study assessing this model empirically (emphasis mine):
Money is frequently used as a means of “buying” services which would probably not otherwise be rendered. Perhaps, then, the presence of money as an external reward suggests to the subjects that they “should probably not render this activity without pay,” that is, they should not be so intrinsically motivated to do the activity.
This could lead the subjects to a process of cognitive reevaluation of the activity from one which is intrinsically motivated to one which is motivated by the anticipation of money.
In this study, Deci paid all subjects for participation in a psychological experiment involving solving puzzles and IQ test questions. Half were notified they would be compensated just for showing up. The other half of subjects were informed that they would be paid upon completion of tasks.
After the presumptive experiment was over, subjects were left in the experimental room with free time during which they could either sit idly or complete more tasks to pass the time. The researchers went behind a two-way mirror and measured the number of IQ questions or puzzles completed during this non-compensated time. They also collected participants’ self-reported measurements of interest in the task during the debrief.
Deci reported that, compared to those who were paid for showing up, subjects who were told they would be compensated upon completing tasks were significantly less likely to complete additional non-compensated tasks and gave lower ratings of interest in the puzzles themselves. Deci interpreted these finding as indicative that intrinsic motivation for and interest in the problems and questions had been displaced by the external incentives (i.e., payment).
Deci’s group followed up on this research the next year. They found that verbal reinforcement by praise/positive feedback did not cause college-aged participants to dislike tasks and complete fewer of them the same way tangible rewards (e.g., money) did; instead, providing positive verbal feedback/praise increased performance and intrinsic motivation.
Another group found that high school students that were promised extrinsic incentives before engaging in a variety of tasks showed less creativity and reported enjoying the assignment less compared to those who were not promised payment at the outset but were compensated at the end unannounced.
In another study, elementary school-aged children 5, 8, and 11 years old were given the opportunity to complete tasks that they were intrinsically motivated to participate in (in this case because the activity was fun and the students were able to choose it for themselves before the experiment). Half of the students were shown a box containing rewards available upon completion of the activity. Half received no mention of reward. As would be predicted by the overjustification hypothesis, students that had been notified of a reward showed reduced free play, and both student and teacher measures of how much “fun” the child had during the activity were markedly lower than children that had not received mention of reward.
In a study of 6-year-olds that were intrinsically highly motivated to draw, Lepper et al. (1973) found that children who were told that they would receive a reward in exchange for drawing came to be less interested in drawing after the reinforcement was given. A group of children was told they would be rewarded for coloring and sketching, and the amount they were engaged in this activity was compared to those who received the same reward unexpectedly and to a group of children that received no reinforcement at all — the kids that were promised a reward for drawing engaged in the task 50% less than those that were unrewarded and those that received a reward unexpectedly.
A more recent study found that, with 3-year old children, there is a similar effect. Letting children know they would receive a tangible reward for sharing served to reduce sharing in the future. Giving verbal praise to these children and providing no reward did not alter the probability the child would share again in the future. In this case, setting up an environment that promised a reward for sharing resulted in less sharing. And this was in 3-year old kids! This adverse effect of external reward on internal motivation seems to be already present before kids even enter pre-school!
I need to stop here for a moment to interject a quick aside before everyone cheers about how verbal praise worked when tangible rewards failed. It is easy to over-interpret this to mean that verbal reinforcement comprises a behavioral panacea for classroom problems. It doesn’t. What the results of these studies indicate is that verbal praise just does not result in the overjustification effect–in college students.
Behavioral research shows that verbal reinforcement by praise is maximally effective when used with college students and becomes increasingly less influential and motivating the younger the child gets. In other words, verbal praise is much more effective with older kids than for younger kids: Our kindergarten students are not as responsive to compliments and praise as high school or college students. I interpret this effect as meaning verbal praise is maximally useful once a child or adult has a well developed intrinsic motivation system in place.
DIRECT IMPLICATIONS FOR BEHAVIORAL MANAGEMENT
The potential negative implications for any behavioral management systems based on incentives and rewards should already be evident. If not, here is a direct quote from the discussion of Lepper’s study in 1973 (emphasis mine):
As an empirical proposition, the present findings have important practical implications for situations in which extrinsic incentives are used to enhance or maintain children’s interest in activities of some initial interest to the child. Such situations, we would suggest, frequently occur in traditional classrooms where systems of extrinsic rewards— whether grades, gold stars, or the awarding of special privileges—are applied as a matter of course to an entire class of children.
Many of the activities we ask children to attempt in school, in fact, are of intrinsic interest to at least some of the children; one effect of presenting these activities within a system of extrinsic incentives, the present study suggests, is to undermine the intrinsic interest in these activities of at least those children who had some interest to begin with. The quite limited manipulation employed in this study, involving a symbolic reward not unlike those routinely employed in the classroom, was sufficient to produce significant differences in the children’s subsequent behavior in a natural preschool classroom. This is consistent with the complaint, from Dewey (1900) and Whitehead (1929) up to the time of Holt (1964) and Silberman (1970), that a central problem with our educational system is its inability to preserve the intrinsic interest in learning and exploration that the child seems to possess when he first enters school. Instead, these authors have suggested, the schooling process seems almost to undermine children’s spontaneous interest in the process of learning itself.
I think these authors make a very salient point on how even 6-year-old preschool-kindergarten kids can have their intrinsic motivation decimated by external rewards in the space of only a single experiment. Remember, this quote is from 1973. This is not a new idea. It has just either actively or naïvely not been taken into account by researchers designing any number of reward or incentive-based behavioral management systems.
HOW SCHOOLWIDE INCENTIVE SYSTEMS REDUCE INTRINSIC MOTIVATION
If you are still asking what this overjustification stuff has to do with reward-based management systems like PBIS, I am happy you asked. These systems are frameworks explicitly designed to use external incentives to motivate prosocial behavior. They were designed to help students with behavioral disorders understand how appropriate behavior is rewarded by having these students watch their peers receive rewards for being “good” and following directions. As such, successful implementations of an incentive program depend on rewarding the good kids (i.e., those that already do the right thing because they are intrinsically motivated) with substantial external incentives for doing the right thing so the other kids can witness them receiving rewards.
The naughtier kids that lack intrinsic motivation to “be good”, it is reasoned, will work to improve their behavior so they also will have access to these large incentives. In most systems, the importance and value of these schoolwide incentives are continuously reinforced in the minds of the students by active schoolwide and classroom economies that are most often used to “buy” tangible (often candy and/or toy-based) rewards and access to activities. These rewards are always available on reward menus clearly and conspicuously posted in classrooms and hallways. In other words: highly desireable rewards are advertised, and access to these rewards are conditional upon completion of certain tasks or norms.
As mentioned above, this design precisely replicates the type of extrinsic motivation that has been shown repeatedly to reduce intrinsic motivation for behaving a certain way or participating in tasks. This is concerning. A framework such as this is great for individuals with behavioral disorders as they, by definition, lack intrinsic motivation to follow the rules and complete tasks; however, a system like this is often problematical for kids that already have developed internal motivation to follow the rules and complete work. Their internal motivation will be challenged within a system such as this. This system tells them they should expect that their intrinsically motivated compliance should be tangibly rewarded.
Researchers have described specific conditions under which extrinsic motivation (like reward incentives, token boards, classroom economies, wall clip charts, etc.) degrades intrinsic motivation and result in students showing less positive behavior than they were intrinsically motivated to in the first place.
These conditions are below:
REWARDS ARE OFFERED IN THE CONTEXT OF PRE-EXISTING INTRINSIC MOTIVATION
This means a reward is offered to a child who is already willing to perform a task for the sake of doing the work. They also might be doing it because they perceive themselves as a “good kid” and they feel an expectation that doing the task is part of being good. Promising or announcing intention to reward kids in this situation does not reinforce their intrinsic motivation; instead informs them that they are to expect an external reinforcer for task completion. Research shows that in these situations, any excitement and anticipation is for the eventual reward, not the task at hand: this negatively affects the quality of task performance.
Reward temporarily increases the student’s probability of performing the task when the reinforcement is present but leaves them unsure how to act in the absence of the reward or when the person that provides the reward is not watching. What was once a task done for its own sake, now becomes the focus of attention and anxiety.
There is now an external pressure to be seen performing to succeed. This kind of anxiety and learning never go well together.
REWARDS ARE KNOWN IN ADVANCE AND EXPECTED
I have no problem with teachers being transparent with students. I feel it is essential to a well working classroom. If there are external incentives, the contingencies for those rewards need to be written in stone and posted (as I have stated before). However, I do not feel a need to post reward menus and have overt displays of available reinforcers in the classroom. This builds an expectation in students regarding what stuff they will get when they comply.
This contract most often serves to motivate students to do work to get the desired reward. If there is no longer the probability of getting a reward, many kids will go into a helpless state and will float and do the absolute minimum to not get in trouble. Similarly, they will avoid doing work on their own and save it up until they are being watched by the person empowered to provide reinforcement.
REWARDS ARE TANGIBLE
This one is huge. Young children have a tough time using only verbal praise or other intangible rewards to motivate behavior. As we get older and gain experience by building our intrinsic motivation, we can more easily accept the words and emotions of others as motivators. That is why high school and college students have the most substantial effects for verbal praise.
Students lose the ability to demonstrate active engagement without access to tangible rewards. This snippet from Lepper’s Discussion discusses challenges with applying token and other external contingent reward systems (emphasis mine)
extrinsic incentives may often be used effectively to increase interest in certain broad classes of activities. On the present line of reasoning, this proposition should be particularly true when (a) the level of initial intrinsic interest in the activity is very low, and some extrinsic device is essential for producing involvement with the activity; or (b) the activity is one whose attractiveness becomes apparent only through engaging in it for a long time or only after some minimal level of mastery has been attained. In fact, such conditions characterize the prototypical token-economy program in that tangible extrinsic rewards are necessary to elicit the desired behavior. Hence, it would be a mistaken overgeneralization from the present study to proscribe broadly the use of token-economy programs to modify children’s behavior.
Some thoughtful proponents of token economies have already recommended that their use be limited to circumstances in which less powerful techniques have been tried and found inadequate (O’Leary, Poulos, & Devine, 1972)—in other words, only when they are necessary. It has also been stressed that in any case, the successful implementation of powerful reinforcement systems demands considerable sensitivity as well as ingenuity on the part of the practitioner (Bandura, 1969). The present study provides empirical evidence of an undesirable consequence of the unnecessary use of extrinsic rewards, supporting the case for the exercise of discretion in their application (O’Leary & Drabman, 1971).
What this suggests is that rather than setting up classwide reward systems, we set up a classroom culture of discovery, and a safely stress-free environment where everyone is free to try new things. If a given student specifically needs an external incentive to motivate their behavior — then we give it to that student because they need it.
Tangible rewards are unnecessary well over 99% of the time. Unfortunately, we are taught in our teaching programs that offering tangible rewards is critical in classrooms and often these are the first thing teachers resort to in a classroom (see here). As a note, I 100% do intend to say this regarding special education classrooms as well — especially those designed for autistic students
REWARDS ARE CONDITIONAL ON TASK ENGAGEMENT
I always tell teachers we need to reward on-task behavior rather than the finished product (“You are working so hard” vs. “You are so smart” or “You got all of them correct”). But there is a way to do this correctly, and a way not to. Sadly, we are all taught the wrong way. If we are to do this well, we need to validate and build upon the student’s intrinsic motivation for dong the work, not by promising them an external motivation if they choose to do so.
How many of us have heard the following: “I would like to give Sven a skittle, but he is not working. Too bad. So sad. Maybe next time. Ebba is working on her spelling, so I am going to give her a skittle for hard work.” I know a lot of students respond just fine to this. Many do not. When paraeducators use this approach is when I most hear kids scream an obscenity and then state that they do not want to be manipulated.
THE RECIPIENT INTERPRETS REWARDS OR OBSERVERS AS CONTROLLING
To be clear, if I give reinforcement contingent on a particular behavior, I AM MANIPULATING THE CHILD. 100%. I AM COERCING them to do what I WANT THEM TO DO if they want reinforcement. And remember, reinforcement is defined as something highly desirable to the recipient. Am I forcing them? No–but I might as well be. I say this because any statement I make negating their access to the reward or putting the reinforcer out of their reach is, by definition, a negative punishment.
Rewards ARE COERCIVE. They ARE MANIPULATIVE. They ARE CONTROLLING.
Giving contingent rewards is not compassionate, kind, or a loving action. Kids understand this fact, and they fight against it. So when a child accuses me of manipulating them, they are right. I usually ask them to explain why I have to manipulate them to be good. I would rather they just did it on their own.
Adults tend to be slow to the uptake. Either that or we have been conditioned over hundreds of thousands of trials to comply, and we have brainwashed ourselves into thinking external rewards provide intrinsic motivation.
How many of us have heard the following: “I would like to give Sven a skittle, but he is not working. Too bad. So sad. Maybe next time. Ebba is working on her spelling, so I am going to give her a skittle for hard work” It is hard for me to imagine a more coercive or controlling situation than using proximity praise and tangible rewards this way. (Note, even giving “stars” for on-task behavior falls under this)
Students learn to perceive external reward as a method teachers use to control them. Invariably, this perception is correct and supported by the child’s daily experience. Students notice whether the teacher is in a good or bad mood and adjust their behavior accordingly.
Worse, many students may learn that they do not deserve the reward because they are not earning any (often these are quiet, well-behaved kids that get overlooked). So they lose both intrinsic motivation and self-esteem.
So what we can see from this post is that this overjustification effect is already established in three-year-olds! Furthermore, it is still present in college students and can negatively impact attitudes toward tasks and reduce task completion in as little as a single session. Clearly, this motivational crowding wherein external motivation displaces and negates internal motivation is a powerful psychological phenomenon that we should be taking into account when we are approaching schoolwide and classwide management strategies.
There are a time and a place for rewards in a classroom. There is a time for external motivation to guide task completion. The time for these approaches is when a given student lacks sufficient intrinsic motivation to perform a task. Only then is it acceptable, and recommended, to provide reinforcement to that child to motivate their behavior. However, we must be willing to accept what we are doing to the child. We are manipulating them to do what we want for payment. This is not healthy in the long term and can irreparably damage the child’s intrinsic motivation system. It needs to be a last, rather than first, option.
I have been through my approach before. I will refer you here and here to learn what I recommend regarding PBIS and here for behavior management in general. Overall, what I wholeheartedly believe is that we need to stop using external motivation as a way of getting kids to engage. We are depriving them of learning for themselves how to act and behave because they want to be good and because they like how they feel when they do the right thing. We are teaching them that, at least in school, their primary motivation to complete work should be to receive a reward from teachers and other adults. We teach them to distrust their intrinsic motivations and desires. We are robbing them of the ability to develop their socioemotional sense of self on their terms.
If we do not help children learn to be self-motivated and to do the right thing because they want to when they are in our K-12 classrooms; then when are they going to learn? College? When they get fired from a job for poor performance? In prison?
Let’s give our students a chance to build character. We can help them build intrinsic motivation without having to resort to empty praise or tangible rewards. We can respect them. Let them grow. Let them mature.