Jody Stallings, a middle school teacher and contributor to the Moultrie News, recently wrote an article titled “Teacher to Parent – Positive reinforcement doesn’t work in the long run” in which he responded to the following question from a third-grade parent:
My third grade son recently came home in tears saying he didn’t want to go to school anymore because he was punished for talking during silent reading. The teacher kept him in from recess. I think this is horrible. It isn’t a teacher’s job to destroy a child’s love for school. Instead of constant punishment for every little infraction, what about using positive reinforcement?
The author dismisses the parent’s suggestion and suggests two reasons he believes positive reinforcement does not work:
- Rewards for good behavior can’t keep pace with children’s changing desires
- Schools shouldn’t prepare kids for a world that doesn’t exist.
On the first point that rewards cannot keep pace, Stallings suggests that rewards that work today will not work tomorrow. The reward will eventually lose its effect, and it will take a greater reward to elicit the desired behavior. Further, he suggests that rewards in middle and high school are all but useless.
For the second point that the real world does not necessarily reward people for doing the right thing, Stallings gives the example of going the speed limit. He poses the question, “Does a police officer pull you over to give you a reward for going the speed limit?”
Stallings hopes to hit home a few points about children. First, students should learn to do the right thing because it is the right thing to do, not because of a reward. Second, actions have consequences, and students need to learn that. Finally, students also need to take ownership of their mistakes and learn to accept the consequences, not have their “love of school destroyed.”
As one would expect with such a polarized topic, the response section got heated. However, one of the biggest contentions with the article was not the positive v.s. negative reinforcement, but the use of recess as a punishment.
Marcy Tanter writes:
It’s appropriate for the teacher to do something, but recess is something kids need to let off steam, which may be why the kid was talking in the first place.
Clisby Williams agrees:
It isn’t that there should be no negative consequence for talking during silent reading. It’s that it should NEVER be removing recess. I don’t know the situation at this child’s school, but 3rd graders really should have 3 recess periods a day: mid-morning, lunch, mid-afternoon. Unless the child’s misbehavior is directly linked to recess (e.g., the child punches other children every time he/she sets foot outside with them) then recess should be off limits. Why not remove the child from the classroom so he has no one to talk to?
In argument to the author’s first point, commenters have pointed out flaws in the author’s suggestions that rewards do not work with older people. Sarah Lanier suggests as much by sharing:
Older kids love no homework passes and extra credit. Those are free. You just have to find a currency that works for the age group.
Sue Conklin continues this point by linking it to the adult world:
In order for Positive Reinforcement to work the ‘reward’ must be something that the learner wants. One would probably not go to work if you were paid in peppermints. Getting paid with money is a form of positive reinforcement. We work for it because it is valuable to us.
On the author’s second point, readers took contention with his assertion of a “real world.” Pam Clark questions this point by writing:
These children are already in the “real world”. They just are not adults yet in the adult world.
Between the article and the comment section, there is a lot to digest, and unfortunately, I cannot address all of it. In order to focus this post, I will address the following questions:
- Does positive reinforcement work?
- Does punishment work?
- Should recess be taken away as a punishment?
To help address these questions, I would like to turn to a meta-analysis of 99 studies on interventions to decrease disruptive classroom behavior done by Stage and Quiroz (1997). Below are selections of data from their study:
Effect Size by Intervention
|No. of Effect Sizes||Mean Effect Size||SD|
Effect Size by Consequence
|No. of Effect Sizes||Mean Effect Size||SD|
|Combined Reinforcement + Punishment||12||-0.97||0.89|
|No Immediate Consequence||70||-0.64||0.54|
Q1: Does positive reinforcement work?
The research shows that positive reinforcement does indeed work. Reinforcement was shown to have a -0.86 effect size. Specifically using token economies, also known as rewards, has shown to have an effect size of -0.90. When compared with punishment alone, positive reinforcement seems to be more effective.
Q2: Does punishment work?
Yes, it does. Although I pointed out above that positive reinforcement has a greater effect than punishment alone, punishment is still effective.
Punishment should not be disregarded because it does not hold up to positive reinforcement. Why? Because the two together are significantly more powerful than either are alone.
According to the research, for schools to get the maximum effect, a combination of positive reinforcement and punishment should be used.
Q3: Should recess be taken away as a punishment?
I felt the need to address the major contention of recess being taken away as many commenters seemed quite upset not about using punishment but about using recess as a punishment.
The research suggests that rather than taking the recess away, a more effective intervention may have been using recess to correct the behavior. Exercise programs as an intervention had a -0.72 effect size compared to the -0.58 effect size of a punishment. Asking the student to have an “active punishment” during recess may have produced better results than taking away the recess.
Based on the research above, I find that the author is generally wrong in his assertion that positive reinforcement does not work. Where I think Stallings goes astray, aside from completely ignoring research, is suggesting that positive reinforcement exclusively equals rewards. While that is one aspect, I think it is more complex than that. He is right that rewards would need to change as the students grow, but I posit that positive reinforcement when you get older does not have to be the trinkets of your younger years.
However, I would also caution about using no-homework passes as rewards as suggested by a commenter. You should have a reason for giving homework to students. It should serve a purpose. If a purpose exists, you wouldn’t want it to be optional. Also, extra credit would be a dangerous reward if you are using grades as data points, which one would hope teachers would.
On Stallings’s second point about “the real world” and rewards, performance bonuses are quite common in many fields and are starting to pop up in education, for better or worse. It seems rewards do still exist even in adult life.
Where I do align with Stallings is that students should learn “bad actions have bad consequences.” As suggested by the research, positive reinforcement alone is not the most effective method. It is best when combined with punishment.
The question then is not whether to use positive reinforcement or punishment, rather it is how do teachers and schools strike a balance between the two.
Stage, S. A., & Quiroz, D. R. (1997). A meta-analysis of interventions to decrease disruptive classroom behavior in public education settings. School Psychology Review, 26(3), 333-368.