Sherman Dorn rounds up the commentary on the recent plans to "pay" students for performing well at school.
None of the commenters, including Dorn, appear to understand "student motivation" and how "rewards" affect motivation or lack thereof. The result is a lot of ill-informed opinion, i.e., the typical state of affairs in education opinion journalism.
As luck would have it, Sherman Dorn showed up in the comments to defend his analysis.
Let me take this opportunity to elaborate on my initial criticism. I'll begin by pointing out that I'm not taking a position with respect to the efficacy of any of the proposed plans. I'm merely criticizing the analyses as being superficial and/or ill-informed. I'll address half of Sherman's analysis in this post and tackle the other half in an other post.
Sherman has two main criticisms of paying for student performance. The first criticism is:
[One] problem with paying students cash for achievement, and that is the question of the reward itself: will it promote continued effort, or will it be tangential to effort?
Apparently, what Sherman means by "tangential to effort" is that rewards are appropriate only "when their use is intimately tied to additional effort."
This definition is unnecessarily limiting.
For a stimulus (such as paying cash) to be a positive reinforcer (or a reward) for a student it must increase the future occurrence of the behavior that will follow. If a stimulus doesn't increase the future occurrence of the desired behavior, it simply is not rewarding or reinforcing for the student. The "tangentialness" of the stimuli is largely irrelevant.
Let's say we have a classroom full of elementary school students. Some like to read, others do not. The behavior we would like see increase is the incidence of reading. We would like all the children to read more because practicing reading helps one become a better reader. To achieve this, we want to find stimuli that will exert some control over each student and motivate them to read more. Stimuli that actually increase the desired behavior (in this case reading) are called positive reinforcers.
There are three main categories of reinforcers that are typically available to a classroom teacher:
- Social reinforcers: involve the student’s behavior—words of praise, attention, smiles, nearness.
- Token reinforcers: are things such as money, poker chips, points, and gold stars that can be exchanged for other reinforcers.
- Activity reinforcers: are activities children like to participate in when given a chance. These might include running, games, art activities, singing, eating, recesses, going home.
For a stimuli to be a reinforcer it must serve to increase the desired behavior. If an an atta-boy from the teacher (social reinforcer) motivates the child to read more it is a reinforcer. If the attaboy does not increase the behavior, it is not a reinforcer. For this child a piece of candy (token reinforcer) might serve as a motivator. For other children additional recess time or a trip to the library (Sherman's example) (activity reinforcers) might be motivators. All the stimuli I've listed might validly serve to promote the "continued effort" of reading.
The flaw in Sherman's analysis is that for most children the trip to the library will not serve as a positive reinforcer because the child most likely already enjoys engaging in the desired behavior, i.e., reading, already. Children who don't like to read are likely not going to find going to the library to be rewarding. And, for those children that do like to read, taking a trip to the library, ostensibly to get more books to read, probably is not increasing the future occurence of the behavior (assuming that the child has sufficient reading material available to him).
I suppose trips to the library might, in very rare cases, serve as a positive reinforcer for a child. Perhaps the child is a connoisseur of library architecture. Perhaps the child loves taking trips, even to libraries, Or perhaps the child is a fan of the Dewey Decimal system. But, I don't think this is what Sherman had in mind.
This also highlights the large catch-22 in Sherman's analysis; only students who already find reading to be enjoyable are likely to be motivated by stimuli "intimately tied" to the activity of reading. Sherman's "intimately tied" limitation effectively forecloses our ability to find reinforcers for students who do not find the underlying behavior we are seeking to increase to be rewarding--the children who most need the motivation.
This flawed analysis leads to a spurious conclusion:
We know that Pizza Hut is engaged in marketing rather than a promotion of reading because it rewards kids with pizza instead of with books.
Both pizza and books are neutral stimuli. Handing a kid a stack of books will not necessarily motivate him to read those books, especially if the child does not find reading to be rewarding already. Unread books do not promote reading.
Pizza, on the other hand, can be token reinforcer for a child who already likes engaging in the activity of, say, eating pizza. The child may be motivated to read a book for the reward of getting a pizza to eat. This would promote reading because the child has increased the amount of books he has read in order to get the reward, the pizza. In fact, operant psychology holds that if a learning activity (reading) is followed by a known reinforcer (pizza), the learning activity (reading) will become a conditioned reinforcer (reading for its own sake). To make the transition it is only necessary to use appropriate procedures to fade out the special reinforcement system.
And, the fact that Pizza Hut is engaging in marketing (its own self-interest) doesn't change this analysis at all. As Adam Smith so famously put it:
It is not from the benevolence of the butcher, the brewer, or the baker, that we expect our dinner, but from their regard to their own interest. We address ourselves, not to their humanity but to their self-love, and never talk to them of our own necessities but of their advantages. Nobody but a beggar chuses to depend chiefly upon the benevolence of his fellow-citizens.
Sherman is mistaken. Rewards/reinforcers do not have to be intimately tied to the underlying desired behavior for them to be effective. To be a reward/reinforcer the only condition is that the reward/reinfocer actually increase or promote the desired behavior.
(Continued in Part Two.)
How do you feel about NYC's pilot program that pays certain low-income familes for doing certian things like taking their kids to the doctor? The program also pays kids for performing well on tests.
Here's are two links with some info:
The (short) NY Post article hints at a future in which tax dollars are used; currently private funds are being used. If families change their behavior for the better, would you be in favor of public financing of such a program?
I've just finished reading Karen Pryor's Don't Shoot the Dog which is life-altering -- and I say that as a person who's spent quite a bit of time in parent-training sessions learning the principles of behavior management.
So, number one: everyone should go read Pryor.
Number two: I'm going to guess that a behaviorist would have the following thoughts concerning money-for-grades.
First: as Ken says, reinforcers increase the likelihood of the reinforced behavior occurring in the future. If money increases the likelihood of students studying more, learning more, getting better grades, etc. then it's a reinforcer & that's a good thing.
Second: Pryor distinguishes between "rewards" and "reinforcers." Things can be rewarding without being reinforcing. We enjoy rewards, but rewards don't make us more likely to do whatever it is we were doing before we got the reward.
Assuming money is a reinforcer for getting good grades, not just a reward, the next question is, "Will students learn to get good grades only for money?"
I think behaviorists would say that this outcome is a distinct possibility. It's called "putting a behavior on cue," and it's a good way to get rid of behaviors you don't want.
To "untrain" a behavior by putting it on cue, you train an animal (or a person) to perform that behavior in response to a cue. Then you never give the cue.
I think behaviorists would say this is a potential danger of money-for-grades. (I realize I'm mixing up "cue" with "reinforcer" here but I think the principle holds...)
Assuming that money works as a reinforcer for good grades, the administrators in charge of this program will need a plan in place to teach kids to get good grades for other reinforcers besides money. I imagine the hope is that getting good grades will become self-reinforcing in time, which is entirely possible.
I don't know whether the NYC program has a serious behavioral component. It was created by an economist.
If the program is being run by administrators following a plan devised by an economist.... then I'm concerned.
Your description of classic behaviorism is tied to a model that is highly individualized: find out what a specific person finds reinforcing, and then tie behaviors to those reinforcers.
There's nothing individualized about either NYC's program or Pizza Hut's, and I'm not convinced that either hypotheses or evidence about the customizing of reinforcers applies to these broad-brush policies.
Sherman, I'm not defending either of these programs. I agree that criticizing the programs as being too broad is a valid criticism. I am merely criticising some of your other arguments which I think are a little off base for the reasons I've given.
NYMT, these programs are problematic because once government starts funding, you will surely get a boatload of regulations, a bureacracy to enforce the regulations, and a new special interest to agitate for more funding.
What happens when you take the pizza away as a reward/reinforcer? Will the kids who previously weren't readers keep reading? Will kids who were readers read more? Less? The same?
The answer is "it depends."
It depends on a few things, all of which should have been asked and answered before the special token system was put into place.
I'll post about this later.
Post a Comment