Wednesday, March 17, 2010

The Case Against Rewards

I had the opportunity to speak at Red Deer College where I gave the case against the use of rewards. I also had the opportunity to sit in and listen to the presenter who provided the case for rewards.

After receiving permission from the speaker to sit in, I knew that I would have to sit quietly and say nothing.

To say that my ears were bleeding by the time the hour was up would be a gross understatement.

In order to reconcile with my bleeding ears, I have decided to provide a rebuttal here on my blog. I will use a Q & A format to provide my ideas.

Don't we all work for pay? How many adults would continue with their jobs if they didn't get paid?

Yes, we do work for pay. I teach middle school and I get paid, and while it is true that I couldn't survive without that pay cheque, there is a big difference between someone who gets paid and someone who works only because they get paid. Many good people who love what they do, continue to do what they love even when financial rewards are removed.

The best employees, the best teachers, the best athletes, the best learners are those who do what they do because they love it. Sure, they might get paid for it, but it isn't nor should it be the driving force in their lives. Everyone needs to make a living. Everyone should be paid fairly for whay they do. And then we need to do everything in our power to take money off the table, to get it out of people's faces.

Studies have shown that things like happiness, type of work and the feeling of making meaningful progress consistantly rank higher than pay. And yet when those employees are asked what they though was important to others, most people said pay.

If you were a principal of a school and you had two teachers of equal value but one was clearly motivated more by the money than the act of teaching, which would you pick? Are the best teachers teaching for the money?

The role of money in the context of work is less prominent than we have assumed.

How do we define the word 'reward'?

To define a reward simply as 'getting something in return' is not accurate. A reward, or as B.F Skinner preferred to say, a reinforcement, is a part of operant conditioning - meaning that an action may be controlled by a stimulus that comes after rather than before.

Some people become distracted by the actual reward itself. I have nothing against candy, stickers, money or smiles. In fact, these can be beautiful gifts that can be bestowed upon those we care the most about. What bothers me greatly is how these things can be used conditionally. To offer these things in a conditional or contingent manner in order to obtain compliance is, ironically, far more heartless than the classroom I promote where the teacher would give candy, stickers and smiles out unconditionally.

Doling out rewards and bestwowing a gift upon someone are two very different things. One is conditional and unloving, while the other is unconditional and loving.

For more on defining rewards, and how the real problem is in the conditional nature of rewarding, read this interview with Alfie Kohn.

I rewarded my child with a computer if he achieved high marks in school, and now he continues to get high marks in school but I only gave him one computer. This proves to me that the use of rewards can encourage achievement.

The idea behind this comment is a familiar one. Basically, this person is assuming a correlation between the bribe and the student's success. And then assumes again that when weined off the bribe, the student still achieves.

The key word here is assume. Firstly, I am concerned that anyone would use the opportunity to learn as a reward or punishment. If I reward a student with a dictionary, he may become a better speller. Shouldn't I have given him the dictionary unconditionally, because it will help him to learn.

Is it not entirely plausible that the student continued to achieve because he had the computer as a tool to enhance his learning? It is far more reasonable to place this child's success on being afforded the opportunity to use the computer to enhance and support his learning - rather than thinking the parent deserves the credit in using operant conditioning to reinforce the child's behaviour.

Judy Cameron and David Pierce's book Rewards and Intrinsic Motivation: Resolving the Controversy uses scientific studies to show that rewards actually don't reduce people's interest in what they're doing.

Upon closer examination, Alfie Kohn's book Punished by Rewards (p260) offers this examination of Cameron's work :

"Cameron's assertion that rewards are basically innocuous depends on drawing conclusions selectively from the relevant research, omitting other studies, and blurring important distinctions. For example, her own review of the data confirms that when people expect to get a tangible reward for completing a task, they do indeed tend to spend less time on that task later than do people who were never promised a reward. But she is at pains to downplay this finding, preferring instead to emphasize that rewards seem not to be harmful under certain conditions, such as when people aren't expecting to receive them (which isn't terribly surprising).

Cameron also argues that negative effects are limited to tangible rewards, whereas the verbal kind are generally helpful. But the way she arrives at that conclusion is by (a) lumping together studies that define praise in very different ways, (b) failing to include studies that found negative effects of praise, and (c) distorting some of the studies that she does include. For example, she points to an experiment by Ruth Butler as proving that "extrinsic verbal rewards" produce extremely positive effects. But anyone who takes the trouble to look up that study will find that it actually distinguises between "comments"and "praise," finding impressive results from the former but discovering that the latter "did not even maintain initial interest at its baseline levels."

What evidence is there to show that rewards do harm to intrinsic motivation?

In 1999, Edward Deci, Richard Koestner and Richard Ryan conducted a meta-analysis which analyzed 128 experiments found that rewards had a significant negative effect on intrinsic motivation. These effects showed up regardless of age, including pre-school to college, with a wide range of interesting activities and with rewards ranging from dollar bills to marshmallows.

In the end, this new meta-analysis showed that "by far the most detrimental type" of reward was the one given "as a direct function of people's performance."

Because no child is at the top of Maslow's hierarchy, they need to be extrinsically motivated in order to progress to the next level.

There are a couple assumptions being made here. Firstly, intrinsic motivation is desirable but that doesn't necessarily mean that is has to be only found at higher levels of Maslow's hierarchy. Some environments may not meet these needs and fulfill our potential, but intrinsic motivation, seen as a function of these needs, is present from the start.

It is also important to note that there is no proof that human needs can be classified into 5 categories, as Maslow suggests. Even if we could assume our needs do fall into 5 categories, there is no proof that we actually transition from one stage to the next. For more on this, read Alfie Kohn's article on Maslow. These misassumptions can encourage some to wrongly assume that extrinsics are required to move from one stage to the next.


  1. An excellent, comprehensive look at the topic! Well done! After reading, I started to think about the connections between the thinking about rewards and the messages we want to send about learning. If we are providing an extrinsic reward for learning (esp. if that's the ONLY "value" we communicate for the learning through our actions), then aren't we conceding that the learning has no value for that student in and of itself?

  2. Wow - that must have been torture! I'm not sure I would have been able to keep my mouth shut. You should find the presenter's email address and send them a link to this post. Maybe also send a link to whoever organized the presentations, encouraging them to share it with the participants and urge the pro-rewards presenter to offer a rebuttal. What can it hurt?

  3. Scott, what an interesting article! We used to reward our children when they did well at school, usually with some treat. I can see the danger in 'if you do this will do this.' that is bribery.



There was an error in this gadget

Follow by Email