by Alfie Kohn
If CL is something more than casual groupwork, it is something less than altruism, or at least less than the narrow version of altruism that requires an individual to sacrifice his or her own interests.5 In asking children to work together, we are not demanding that they ignore their own academic well-being in order to make sure that someone else understands the material. Positive interdependence means that when you succeed, I succeed, too; my interest in your learning is matched by your interest in mine.
No catalogue of misconceptions about CL would be complete without reference to the idea that working in groups is just a gimmick to perk up a classroom now and then, providing a break from serious instruction. (“Okay, kids, it’s the third Friday of the month. Remember, that means today we work in teams!”) While teachers doubtless will want to continue making some use of whole-class discussion6 and individualized work, CL can—and, I would argue, ought to—become the standard or “default” classroom arrangement starting in kindergarten.
Finally, to echo my earlier remarks on constructive conflict ([>]), CL does not entail harmony or concurrence about controversial issues, even after a discussion has been concluded and an assignment completed. Teachers who properly want to avoid constructing adversarial encounters need not demand a premature consensus: “Your assignment is to come to an agreement about whether abortion should be legal.” One likely result of such an instruction is that the student with the loudest voice, or the one holding the pencil, will write down whatever she wants, everyone else in the group will give in so as to be done with the whole business, and no one will learn anything. Conversely, teachers who properly want to preserve controversy need not set up a win/lose structure to do so. In a classroom whose members have come to feel part of a community and who have been helped to develop a range of social skills, disagreement can occur without debate, and conflict without competition.*
Perhaps it is easier, then, to say what CL is not than to say what it is, at least beyond a one-sentence formula that is about as helpful as saying a poem is a careful arrangement of words. Yes, we can add that in classrooms where CL is used, teachers communicate the message “I want to see what you and your neighbor can do together.” They are likely to ask students to turn first to a peer to check out an idea or request an explanation. They may evaluate some projects as group efforts, which is what they are. These teachers probably have arranged the room so that children are clustered around tables most of the day instead of being seated at separate desks. They recognize that “socializing” is not something relegated to recess and lunch, something that distracts from learning; rather, they know that learning emerges not only from what transpires between student and teacher, or between student and text, but also from what happens between student and student.
Beyond this, though, it is difficult to say with certainty what CL is, for the simple reason that it is strikingly different depending on which theorist, trainer, or teacher you ask. From a distance, the distinction between CL and traditional classroom arrangements is so pronounced that those who promote learning in pairs or groups seem distinguished principally by this commitment. But seen from up close, those who huddle together under this conceptual umbrella often differ from one another in the way they conceive of cooperation and, for that matter, learning itself.
THE EFFECTS OF COOPERATIVE LEARNING
Before exploring some of the different varieties of CL, I want to say why that discussion—and, for that matter, the whole movement—is so terribly important. Several versions of CL have been tested in literally hundreds of studies, and the results of that research should be much better known than they are at present. The gains realized from CL are so impressive and consistent that it is difficult to understand how any talk of school reform could omit the critical question of whether students should be learning alone or together.
SELF-ESTEEM: Consider first the issue of how a student feels about himself and his abilities. The evidence in chapter 5 might be summarized as follows: competition is to self-esteem as sugar is to teeth. Far from providing a lasting sense of confidence in oneself, a competitive environment leads one to ask “What have I done for me lately?” But CL does not merely avoid the abrasive effect of competition: at its best, it has a fortifying effect of its own.
Recent reports of CL’s impact on self-esteem are remarkable. Positive outcomes were found in “eleven of the fifteen studies in which the effects of cooperative learning on self-esteem were studied,” Robert Slavin wrote in 1990.7 In a much larger review of the research comparing competitive and cooperative arrangements on selfesteem—a review that, for better or worse, lumped together studies of many different types and of varying quality—the Johnsons reported that a fair number failed to find any statistically significant difference between the two conditions. But of those studies that did turn up a significant difference, exactly one showed an advantage for competition while eighty-one showed an advantage for cooperation. This sort of consistency is, to put it mildly, very uncommon in the social sciences.8
On the other hand, it should not be concluded that CL is a magical cure for low self-esteem. First, children sometimes feel lousy about themselves because they are abused or humiliated at home or do not get enough to eat; these problems require changes far deeper and wider than bringing CL to the classroom.9 Second, the research on self-esteem (from which conclusions about the effects of educational arrangements are drawn) leaves a great deal to be desired in terms of rigor. While many of us have gotten in the habit of casually tossing the term around, a global version of self-esteem (how you feel about yourself in general) is far less useful, scientifically speaking, than specific aspects of self-esteem (how positive you feel about your capacity to do math problems, for instance). Moreover, the tests used to measure self-esteem are often idiosyncratic and badly designed, which means that the result of one self-esteem experiment cannot meaningfully be compared to the result of another that used a different test.10 Finally, even some of the better pencil-and-paper measures are intended to tap how one feels about oneself at the time the test is taken, and this tells us little about one’s persistent, stable, underlying beliefs or feelings about oneself, assuming such things can be identified or measured at all.
Notwithstanding all of these (and other) limitations on the theory and practice of self-esteem, there is reason to believe that to whatever extent the time a child spends in school can make a difference in how she sees herself, CL is far more likely to promote a positive self-image than are conventional methods of instruction.11 Learning in well-functioning groups provides an environment of social support and enhances the chances of academic success; these effects, in turn, help children to develop more confidence in themselves, more resilience in the face of failure, and more of a sense that their fate is (at least sometimes) in their own hands rather than at the mercy of external forces.
SOCIAL INTERACTION: No one is likely to be astonished by the fact that a structure of positive interdependence (your success equals my success) inclines us to look more favorably on the people with whom we are interacting than a structure of negative interdependence (your success equals my failure) ([>], [>]). The only question for those with an interest in education is whether reducing hostility, improving social skills, promoting an acceptance of people from different backgrounds and with different abilities, and coming to view others as potential collaborators (rather than as obstacles to one’s own success) are judged to be sufficiently important goals to warrant the use of CL. To ask this question is, I think, to answer it. In fact, David Johnson has said that his work is motivated more by a desire to help children accept people who are different from themselves than by a desire to boost achievement.12
Once again, recent reviews of evidence confirm that CL leads children to view one another more favorably, increases the likelihood of cross-ethnic friendships, enhances acceptance of students with handicaps, and promotes perspective taking (that is, the capacity to imagine someone else’s point of view).13 Moreover, “coope
ration seems to promote better relationships when intergroup competition is absent.”14 At a time when managers frequently complain that employees seem unable to get along with each other or work together effectively—and, indeed, “poor interpersonal skills represent the single biggest reason for failure” of the managers themselves, according to one study15—the fact that children learn social skills from CL seems to be a persuasive argument in itself for making American education less competitive and more cooperative.
ACHIEVEMENT: The most gratifying (and perhaps surprising) finding from CL research is the news that there is no trade-off between helping students to feel better about themselves and each other, on the one hand, and helping them to learn more effectively, on the other. Working together offers “bottom-line” benefits in terms of achievement just as surely as it enhances self-esteem and the quality of relationships. These benefits are enjoyed by students of all ages, in all subjects, and at all kinds of schools.
Anyone who finds personal testimonies compelling ought to seek out a teacher who has been using CL over a period of time and ask what the effect has been on students’ comprehension, problemsolving strategies, creativity, or even simple recall of facts. The change can be literally difficult to believe. One teacher in New York told me that her students’ academic performance improved so drastically after they began working in teams that her principal noticed the difference in her grade reports. Rather than asking what accounted for the children’s higher achievement, however, he demanded to know why she had suddenly become an easier grader.
Research confirms the improvement, even on relatively superficial measures like standardized tests. Slavin picked sixty-eight achievement comparisons, making a point of excluding studies he thought were inferior and including a disproportionate number of his own. Of these, forty-nine (nearly three quarters of the total) found superior results from CL as opposed to standard classroom arrangements.16
At the other end of the continuum, methodologically speaking, the Johnsons reviewed every available study from 1898 to 1989—some 369 in all—that examined task achievement in cooperative, competitive, or independent conditions, (Only some of these were conducted in classrooms.) In comparing the effects of cooperation and competition, no significant differences were found about one third of the time. When there was a difference that probably was not due to chance, cooperation produced better results than competition did in 87 percent of the comparisons. Cooperation produced better results than individualized task achievement in almost exactly the same proportion where there were significant differences.17 Elsewhere, the Johnsons have observed:
That working together to achieve a common goal produces higher achievement and greater productivity than does working alone is so well confirmed by so much research that it stands as one of the strongest principles of social and organizational psychology.18
This principle has been noticed even at the farthest reaches of higher education. A 1992 assessment of teaching and learning at Harvard University, based on interviews with 570 undergraduates, led to the central conclusion that intellectual growth was associated primarily with “interpersonal activities with faculty members, or with fellow students, built around substantive, academic work.” The report urged the creation of study groups where they do not already exist and noted that many students avoided taking science classes not because of the heavy workload but because of the competition for grades.19
Chapter 3 offers a few reasons to explain why competition in general proves less successful than cooperation.* Here I would like to summarize and supplement those accounts, widening the focus of our inquiry to ask how it is that CL is so effective at helping children to learn. Some of these reasons are negative: CL works because it avoids having children compete, thus avoiding a number of counterproductive effects.
1. ANXIETY: Competition promotes anxiety of a kind and level that interferes with performance.
2. EXTRINSIC STATUS: Like other extrinsic motivators (that is, artificial incentives outside the task itself) such as money, grades, certificates, extra dessert, and gold stars, competition leads children to complete a task as quickly as possible, which means they avoid taking risks and exploring ideas in the way that is necessary for creative problem solving and deep learning. Competition and other extrinsic motivators tend to reduce interest in, and ultimately impede performance on, many kinds of tasks.20
3. ATTRIBUTIONS: Whether they win or lose, children typically attribute the results of a competitive encounter to luck or fixed ability. The result is a diminished sense of empowerment and responsibility for their learning.21
4. PREDICTABILITY: If the chance to win a contest is the chief reason students have been given for learning something, most of them have figured out in September who is likely to come out on top. The presumptive winner therefore has been given no reason to do anything more than is necessary to defeat everyone else. More important, all the likely losers have been given no rationale for bothering with the subject matter at all.22
The remainder of the reasons are positive: CL works because of benefits that cooperation itself offers, namely:
5. EMOTIONAL BENEFITS: CL’s positive effect on self-esteem and relationships with peers tends to translate into a positive effect on achievement. Our experience in the workplace bears this out: people who feel good about both themselves and their colleagues are more likely, all things being equal, to do a better job. Self-esteem, relationships, and performance are integrally related so that enhancing any one variable can positively affect either of the others.
6. ELIMINATION OF NERD STIGMA: A student in the conventional classroom who gives evidence of being too interested in the lesson is liable to be on the receiving end of any number of unlovely labels that have replaced “teacher’s pet.” But CL helps to create a new set of norms. “When students are working together toward a common goal, academic work becomes an activity valued by peers.”23 CL, in other words, makes it cool to learn, and this increases the probability that effective learning will indeed take place.
7. INTEREST IN THE SUBJECT: Not only is it more fun to be part of an effective and supportive group than to work against, or apart from, others, but CL also leads students to become more enthusiastic about what they are learning. (This is surely an independent benefit: beyond how children score on today’s test, we want them to become hooked on playing with words and numbers and ideas.) The Johnsons cite twenty-two studies to support the claim that CL promotes “more positive attitudes toward the subject area being studied and the instructional experience.”24 One of their favorite stories concerns a teacher who advised them to warn other participants in their training sessions against using CL just before lunch. Puzzled, they asked why. The teacher explained that each class in her school was supposed to arrive in the cafeteria at a different, precisely assigned time in order to prevent long lines from developing. When her students were working in groups, though, it was difficult to drag them away from what they were doing. She would tell them it was time to stop for lunch, and they would protest: “Wait! Wait! We’re almost finished!”*
By contrast, students in most American classrooms are eager to be free of them. They count the minutes until the end of the period, the days left before the weekend, the weeks they must endure until the next vacation. Perhaps this is because they hear their parents articulating similar sentiments about their work by thanking God for Friday. But maybe students genuinely find life at school to be a collection of tedious tasks and humiliating evaluations from which any reasonable person would want to escape. John Goodlad’s mammoth study of more than a thousand representative classrooms across the country confirmed that “the kinds of classroom practices found most often were well liked by relatively small percentages of students”; the older the child, the less satisfaction was expressed.26
Of course, everyone is aware that most kids do not care for most aspects of school. Everyone is also aware that too many students are graduated without the intellectual skills or knowledge t
hat we expect the schools to have provided; indeed, it has become a popular entertainment of late to describe how much students do not know. But rarely do we connect these two pieces of data. If children seem unhappy about going to school, we typically attribute this to the fact that kids just are wont to complain, that they don’t like anything—or at least anything good for them. Then we insist that they had better get used to things that aren’t any fun. (The premise here seems to be that the chief purpose of school is not to get children excited about learning but to get them acclimated to doing mind-numbing chores.)
It is also possible, however, to conclude that the problem may just lie with what happens in school rather than with some character flaw in the individual child. The extent of “on-task” behavior in a classroom, by the same token, may tell us something about the teacher as well as about the students. When a teacher complains that children are off task, our first response should be to ask “What is the task?” (In the long run, though, individual teachers probably are not to blame, given that decisions about what children must learn and how they must learn it are frequently made by administrators, school board members, parents, politicians, and faculty members at schools of education.) Moreover, the sharp line drawn between enjoying the process, on the one hand, and buckling down to learn how to spell, on the other, may reflect a philosophy of self-denial—or, more accurately, other-denial—that yields the worst of both worlds, since it often produces neither enjoyment nor effective results.
