Dr. Spencer Kagan

To cite this article: Kagan, S. Kagan Structures and Learning Together: What is the Difference? San Clemente, CA: Kagan Publishing. Kagan Online Magazine, Summer 2001. www.KaganOnline.com

I am grateful for the very helpful input of Liana Forest and Miguel Kagan who responded to an earlier draft of this manuscript.

I am often asked about the difference between the Kagan Structures and the Learning Together models of cooperative learning. The Learning Together model, developed and advocated by David and Roger Johnson, is a powerful contribution to education both in theory and in practice. The Kagan Structures model has many areas of convergence with the Learning Together model, both in theory and practice. Among the areas of convergence in the two models are belief in the importance of Positive Interdependence, Individual Accountability, Social Skills Development, and Face-to-Face Interaction. (There is a difference in how the two models define positive interdependence, but that difference is rather technical and more of concern to researchers and theoreticians than to most teachers – See Appendix 1.) Both models provide intensive training in their respective differentiated conceptual frameworks and both models call for practice in numerous practical implementation techniques.

Nevertheless, the two models differ in a number of very important ways. The models attribute different relative importance to the elements they have in common, but more importantly, they each promote practices not promoted or even discouraged in the other model. Teachers trained in one model end up teaching lessons which are quite different from those created by teachers using the other model.

To shed light on the most important differences in the two models, I present a thumbnail sketch of the essential elements of the two models, identify and discuss six important differences, and spell out implications for implementation.

I. Essential Elements of the Models – A thumbnail sketch

The Johnsons postulate five conditions under which cooperative learning is productive:

1. Clearly perceived Positive Interdependence

2. Considerable promotive Face-to-Face interaction

3. Clearly perceived Individual Accountability and personal responsibility to achieve the group's goals

4. Frequent use of the relevant Interpersonal and Small-Group Skills

5. Frequent and regular Group Processing of current functioning to improve the group's future effectiveness

Kagan offers six keys to successful cooperative learning:

1. Effective formation and utilization of Teams

2. Development of the Will among students to work together

3. Efficient Management techniques

4. Development and practice of Social Skills among students

5. Appropriate implementation of Structures

6. Inclusion of four basic principles, symbolized by the acronym PIES

II. Six Important Differences

1. Structures

The greatest point of divergence in the two models are Kagan Structures. The Kagan Structures model, as the name indicates, places great emphasis on structures. The Johnsons do not include structures in their model. The Kagan model defines over 150 repeatable, step-by-step, content-free ways to structure the interaction of students with each other, the curriculum, and the teacher. The structures are very empowering for a teacher. For example, once a teacher understands the steps of one simple structure, RallyRobin (Students in pairs take turns saying something), the teacher can use RallyRobin to have young students name colors or older students name prime numbers.

RallyRobin can be used as a set for any lesson (name things you already know about the topic; name things you would like to learn about the topic), for practice (take turns inserting colorful adjectives into a sentence frame), or for closure (name things you have learned).

Once the teacher knows any one structure, the teacher can easily generate an infinite number of activities. The basic formula in the Kagan model is Structure + Content = Activity. The teacher merely "plugs in" his/her content into one of the Kagan structures to create a new, engaging activity for students. Because the structures have the basic principles of cooperative learning "built-in" to their steps, the job of the teacher in creating solid cooperative learning activities is very easy.

These structures have been at the heart of the Kagan model since its inception, both in theory and practice, and are not included in the LT model. Structures have been crafted to deliver differentiated programs in Multiple Intelligences (Kagan, S. & Kagan, M. 1995; Kagan, L., 2000); Character Development (Kagan, S. 2000); Higher-Level Thinking (Kagan, S. 1994); Brain-Compatible Learning (Kagan, S. 2001) and Standards (Kagan ,S., Kagan, M.& Kagan, L., 2000a, b, c, d). Within cooperative learning, distinct structures have been developed to facilitate teambuilding, classbuilding, mastery, thinking, information sharing, and communication skills (Kagan,S. 1994; Kagan, L., Kagan, M. & Kagan, S. , 1997; Kagan, M., Robertson, L. & Kagan, S., 1995). Books demonstrate the power of structures for delivering core academic curriculum at the primary grades (Candler, 1995a; Curran, 1998; 2000), middle school grades (Andrini, 1998; Candler, 1995b; DeBolt, 1998; Stone, 1994) and high school grades (Kushnir, 2001; Morton, 1998). Specific structures have been created to develop specific types of reasoning (Kagan, M., 200; Wiederhold, 1995) and language development (High, 1993). The structures are flexible, powerful tools which make teaching easier and learning more engaging and successful across the range of grades and academic content areas.

2. Lessons v. Structures

The cooperative learning lesson is at the heart of the Johnson model: To quote the Johnsons, "cooperative learning groups have five essential elements (positive interdependence, individual accountability, face-to-face promotive interaction, social skills, and processing) built carefully into every lesson to teach the students to learn well together. Learn how to include them in each cooperative lesson." They call for teachers to "tailor cooperative learning lessons to meet the unique instructional circumstances and needs of the curricula, subject areas, and students."

In contrast, for a number of years I have been telling teachers not to design cooperative learning lessons. Rather, I urge them to make cooperative learning part of every lesson by including structures. This distinction cannot be dismissed as simply a difference in semantics. In the Johnson model, the goal of the teacher is to become competent in planning and delivering cooperative learning lessons; in the Kagan model, the goal of the teacher is to become competent in a range of structures. In one case the focus is on the lesson (which might last a day or so); in the other case the focus is on the structure (which might last from 1 to 15 minutes). A teacher successful in the LT model becomes efficient in planning lessons. A teacher successful in the Kagan model becomes skillful in a range of structures.

In one case, faith is placed in the lesson plan and effort is required to redesign each lesson to include the five essential elements in the LT model. In the other case faith is placed in the teacher learning and using structures. Training in the Johnson model requires practice in lesson planning. Training in the Kagan model requires practice in structures – the steps of the structure and their domain of usefulness (when to use and not use a structure).

Once a teacher obtains a stable repertoire of structures, they can make any lesson a cooperative learning lesson with little or no special planning. The most important goal in the Kagan Structures model is for teachers to reach "unconscious competence" in a range of structures. That is, to know a range of structures so well they can use them skillfully with little or no effort. When we learn to drive a car, at first, we have to pay attention to all the discrete skills (How do I shift gears? How hard do I press on the accelerator? When do I begin to signal a turn?). Later we perform all those skills automatically – we reach unconscious competence. When a teacher reaches unconscious competence in a range of structures, he/she creates engaging, powerful cooperative learning activities as part of any lesson with little or no special planning or effort.

3. Basic Principles

In the Kagan Structures model, four basic principles of cooperative learning are identified, symbolized by the acronym PIES. The first two of these principles, Positive Interdependence and Individual Accountability are also identified in the Johnson model as two of the five conditions of productive cooperative learning (see footnote). There are, however, important differences, in the two models in their treatment of the last two of the PIES principles, Equal Participation and Simultaneous Interaction.

Equal Participation. The third PIES principle is Equal Participation. The Learning Together model does not address structuring the interaction among group members so there is equal participation. This is a major point of divergence. For example, teachers trained in the LT model might be (and often are) satisfied with having students discuss something in a group or as a pair. Those trained in the Kagan model are not satisfied with unstructured Group or Pair Discussion, in part because the Kagan model emphasizes Equal Participation. Usually during unstructured group or pair interaction, there is very unequal participation and those who would most benefit from the opportunity to verbalize their ideas, are least likely to verbalize! Leaving equalization of participation up to the students is wishful thinking and almost always results in unequal participation. Even among socially skilled individuals, participation is usually very unequal unless a structure is used. That is part of the reason structures like Timed Pair Share, RoundRobin, and RallyRobin are so powerful – they ensure each learner verbalizes for approximately the same amount of time.

It is recognized that participation will never be truly equal during cooperative learning because students have unequal gifts and so their contributions differ. With regard to what students contribute the term "equitable" is better than equal. The teacher cannot control how equal the contribution of each student will be, but the teacher can control and structure for more equal opportunities for contributions. For example, if the students are to make a list, the teacher can call for a RoundTable (each in turn writes one answer), rather than allowing one student to write the list for the team. If students are to share their ideas with a partner for two minutes, the teacher cannot predict how equal the contribution of each will be, but the teacher can call for a Timed Pair Share rather than a Pair Discussion, providing each student a minute to share rather than allowing one student to dominate the discussion and take most or even all of the two minutes.

Simultaneous Interaction. The Learning Together model calls for "Face-to-Face" interaction whereas Kagan emphasizes "Simultaneous Interaction." This apparently minor difference has important implications. Kagan defines simultaneous interaction as the percentage of learners overtly engaged at any one moment. This is an important definition because the percentage figure tells us clearly that a group of four will create more interaction than a group of five, and that pair work doubles active participation compared to square work (work in a team of four). There is nothing in the call for "face-to-face" interaction that tells us that a group of four is better (in terms of amount of participation) than a group of five, or that pair work is better than square work. Without calculating the percentage of active participants at any one moment, the teacher might think that a group of five is as good as a group of five for promoting active, engaged interaction. In fact, in the same amount of time in the group of three each student talks almost twice as much as in a group of five.

4. Teams

Teams v. Groups. The Johnsons use the term "Groups;" Kagan uses the term "Teams." In the Kagan Structures model we explicitly distinguish teams from groups: A bunch of strangers standing on the street corner waiting for a stop signal to turn green is a group. For that group to become a team, much more has to happen. Thus there is a great deal of emphasis in the Kagan model on Teambuilding, and structures and activities explicitly designed to accomplish the five aims of teambuilding (see Kagan, Kagan, & Kagan, 1997).

Team Size. An important difference in the two models is team size. The Johnsons recommend groups consisting of two to five members; the Kagan structures model places heavy emphasis on the need to create teams of four as often as possible. The call for teams of four in the Kagan model is based on two of the PIES principles: Equal Participation and Simultaneous Interaction. Equal Participation tells us that during pair work, one person is left out in a group of three or five, so groups of four are preferable. Simultaneous Interaction tells us that groups of four maximize active participation because groups of four break neatly into two pairs, allowing the maximum amount of simultaneous interaction – 50% of the class producing ideas at any one moment. To give each learner a minute to express his/her ideas takes five minutes in a team of five, but only two minutes during pair work. Whereas saving three minutes of time does not seem like much, if the teacher saves just ten minutes of time each day, the teacher has freed up about an hour each week to create additional learning opportunities.

Types of Teams. In Learning Together, three types of groups are recommended: Formal Cooperative Learning Groups (lasting form one class period to several weeks), Informal Groups (lasting from a few minutes to one class period), and Cooperative base groups (lasting a semester or a year). In the Kagan model, a wide range of types of teams are distinguished along with recommendations of when to use each, how long to keep them together, and a variety of efficient ways to form them. Among the types of teams used in the Kagan model are Heterogeneous Teams, Random Teams, Language-Specific Teams, Interest Teams, Jigsaw Expert groups of various types, Student-Selected Teams, Topic-Specific Teams, and Issues-Specific Heterogeneous Teams based on a Folded Agree-Disagree Line-up are all part of the Kagan model.

5. Social Skills

Both models identify social skills development as important because it facilitates successful cooperative learning and because it is an important educational outcome in its own right. There are, however, some important differences in the approach to social skills taken in the two models.

The Johnsons elevate social skills to one of their five essential elements defining cooperative learning. In the Kagan model, social skills are not one of the PIES principles which define cooperative learning, but are treated in depth as one of the six keys. Why is there more emphasis on social skills in the Johnson model? Why must there be a social skills component for each lesson? It turns out that if you don't have structures, you need to take time to assign roles and process social skills. Because social skills are embedded in the Kagan Structures, often there is no need to take extra time for the social skills component of a lesson or activity.

Lets take an example. Students are to make a list in their team. Kagan uses RoundTable, each student in turn adding one item to the list. In the Johnson model, with no structure to rely on, students are told to make a list in their team. To equalize the participation, a role of Gatekeeper is assigned. The Gatekeeper's job is to make sure everyone's participation is about equal. An Encourager might be assigned as well, to help "bring out" the shy or reticent students. Further, following the activity students process, asking how equal their participation was and what they could do to encourage more equal participation. Using the RoundTable structure in place there is no need for the roles or group processing – the structure equalizes the participation; everyone participates about equally. Using Kagan Structures radically reduces the need to assign roles and process group interaction.

Without assigning roles or processing social skills, students acquire social skills if structures are used. We have observed even in Kindergarten classrooms, if RoundTable has been used frequently, students begin to spontaneously RoundTable their work, without the teacher calling for the structure. When taking turns is simply the way we do things, it becomes internalized. Students who have played Paraphrase Passport often, begin to use paraphrasing during an unstructured discussion. A social skills curriculum is embedded in the structures and is acquired without the use of role assignment and/or processing.

This difference is quite important, especially in face of the increasing press to reach high academic standards. In the press to get through the academic content, teachers are likely to find they do not have time to design and implement separate lessons on social skills or to give equal weight to the social skills and academic components of a lesson. They do, however, have time to use structures while delivering the academic curriculum. And because the social skills are embedded in the structures, by using a range of structures, the teacher effectively delivers a social skill curriculum without taking time off the traditional academic content.

Let's take an example. A teacher uses a Team Statement to have students reflect on the meaning of democracy. Each student in turn writes his/her own definition, reads it to the group, receives appreciation from the group, and then the group works to come up with a team definition which captures the essence of the concept better than any of the individual definitions. The teacher is having students construct meaning on an important social studies concept. But at the same time students are learning to seek consensus, turn taking skills, listening skills, praising, supporting, asking for information, giving information, asking for help, and giving help. Using structures the teacher delivers a much richer curriculum. Without taking a minute from the academic lesson, and without special preparation and planning time, the teacher delivers a social skills curriculum.

6. Group Grades

Group grades are defined as each member of a group receiving points or grades based on the achievement of the group as a whole – either test scores above a certain criterion or a group product of a certain quality. I call for teachers to abandon using group grades (Kagan, 1995, 1996). I have argued group grades are

1. Unfair

2. Debase report cards

3. Undermine motivation

4. Communicate to students that their grade is a function of forces beyond their control

5. Violate individual accountability

6. Create resistance to cooperative learning

7. Are, or at least should be, illegal

The Johnsons, in contrast, have at times advocated group grades or bonus points for the group if all members score above a certain level. For example, in their discussion of how to produce positive interdependence, they write,

Positive Reward – Celebrate Interdependence. Each group member receives the same reward when the group achieves its goals. To supplement goal interdependence, teachers may wish to add joint rewards (e.g., if all members of the group score 90% correct or better on the test, each receives 5 bonus points). Sometimes teachers give students: 1) a group grade for the overall production of their group, 2) an individual grade resulting from tests, and 3) bonus points if all members of the group achieve the criterion on tests. (Johnson & Johnson, 2001).

On their Web page question-answer section, they have defended group grades as something a veteran cooperative learning teacher has no difficulty in explaining:

A teacher writes:

I have been teaching (and using co-operative learning) for 10 years. I have always used group marks as part of the process. I recently read an article from Spencer Kagan saying group marks should not be used. What are the benefits and the downside of using group marks from your perspective.

Roger Johnson replies:

The question of grades and cooperative learning always comes up and there are several variables to consider:

Group grades follow group goals and give a strong message of Positive Interdependence. Students who are veterans of cooperative learning prefer group grades as they insist that everyone contributes in a different way and there is no way to sort it out except to share the mark.

Teachers new to cooperative learning sometimes are challenged before they are ready to defend the use of cooperative learning if they give group grades so we usually advise rookies to not give group marks on serious material instead to prepare each other on the test and project and take the test or defend the project alone. This is not a bad procedure as individual accountability is part of the cooperative learning group (learn it in the group and be able to demonstrate it alone). On the other hand our veterans (and you would fit that category by the sound of it) use group grades whenever it is more appropriate to do so and have no difficulties explaining it to questioners.

I consider myself a veteran cooperative learning teacher and find group grades not only difficult to defend but totally indefensible and, frankly abhorrent. With group grades in place, students who do little or nothing sometimes get a free ride; students who do exceptional work sometimes have their grades lowered by being placed on a team with an unmotivated teammate. It is unfair to reward (or punish) students for the work (or lack of work) of others. With group grades in place students who are identical in motivation, ability, productivity, and learning can receive different grades.

While the Johnsons have advocated and defended group grades, they do caution readers about their use: In a list of "Common Mistakes In Using Cooperative Learning - And What To Do About Them!" they include group grades. But even there they indicate group grades are sometimes acceptable, stating, "Give group grades only when absolutely necessary, absolutely fair for each member, and when you have taught the students how to work together." My response: group grades are always unfair because they base a student's grade in part on the performance of other students.

III. Implications for Implementation

There are many differences in actual classroom practices of teachers versed in the Kagan Structures and those versed in the Learning Together model. In what follows I will focus on two core issues: 1. Moment-to-Moment Interactions, and 2. Sustained Implementation.

Moment-to-Moment Interactions

What happens in a classroom on a moment-to-moment basis determines, to a tremendous extent, the educational outcomes for students. All of those moments add up. For example, if a teacher has students interact with a partner on the average once an hour, and the interaction is marked by the higher achiever doing most or even all of the talking, then there are five lost learning opportunities for the lower achiever each day, or an average of of 25 lost learning opportunities a school week! What we do each time we have students interact is terribly important, because the sum of all those interactions determines, to a tremendous extent, the learning outcomes in our classrooms. Kagan Structures make it easy and efficient to maximize the learning potential in all those moment-to-moment interactions whereas the Learning Together model does not handle those mini-cooperative learning moments efficiently. To make that case, let's examine two ways to structure a few minutes of interaction.

Turn to a Partner v. Timed Pair Share

As indicated, to create classroom applications which respect the basic definitions of cooperative learning in their respective models, Kagan has relied heavily on structures and the Johnsons have relied heavily on group processing and teaching social skills. This results in classrooms which look and feel quite different.

Let's analyze a very simple example. A teacher has been lecturing and wants students to interact for a few minutes over an issue highlighted in the lecture. The lecture and subsequent discussion might be on any topic (the assumptions underlying the concept of manifest destiny, the conflicting motivations of a character in literature, alternative explanations of a puzzling science experiment, or the pros and cons of two alternative algorithms). Following a Kagan training, the teacher might have the students interact using one of the Kagan Structures, say, Timed Pair Share. In a Timed Pair Share, students are in pairs, Student A and Student B. "A" in each pair speaks for a specified time, receives feedback from his/her peer, and then "B" does the same. Timed Pair Share, like all of the Kagan Structures is carefully designed to include the PIES principles:

PIES Present in Timed Pair Share

Positive Interdependence: + The ideas of one student enrich the thinking of the other; each must contribute.
Individual Accountability: + Each is required to perform in front of a peer.
Equal Participation: + Each performs for the same amount of time.
Simultaneous Interaction: + Half the class is verbalizing ideas at any one moment.

Using a Timed Pair Share the teacher is confident that good cooperative learning has occurred as defined in the Kagan model because the PIES principles are "built into" the structure.

Without training in the structures, a teacher might well say "Turn to a partner, and talk over the issue." The teacher is using unstructured interaction. Pair Discussion is not crafted to contain all of the PIES principles.

PIES Missing in Turn to a Partner

Positive Interdependence: + The ideas of one student enrich the thinking of the other.
Individual Accountability: - Students are not individually accountable for talking.
Equal Participation: - One student may do most or even all of the talking.
Simultaneous Interaction: + Half the class is verbalizing ideas at any one moment.

Now let's assume for a moment the teacher is trying to apply the LT model and wants to convert the interaction from unstructured interaction to solid cooperative learning to meet the criteria of cooperative learning in the Learning Together model. The teacher's job is to work from the five basic principles to create a cooperative activity. Working from the LT principles only, to align "Turn to a Partner" with good cooperative learning, the teacher would have to define a social skill and use group processing on the social skill. For example, if the teacher wanted to focus on equal participation the teacher would have students take time to ask how equal their participation had been, give each other feedback on how equal it had been, celebrate if it had been equal, and make a plan to make it more equal next time. The plan might include assigning the role of Gatekeeper, whose job it is to equalize participation. That is a pretty cumbersome sequence of events to equalize participation in a few minute interaction!

The teacher using Kagan structures is assured that good cooperative learning has occurred and is free to move on to creating an additional learning opportunity while the teacher working in the LT model is burdened with defining and processing a social skill. The structures are an efficient solution to converting unstructured interaction into solid cooperative learning, especially in the moment-to-moment interactions which should occur on a frequent basis if we are to reap the benefits of cooperative learning.

This is not to say that processing and teaching social skills are bad. On the contrary, they are many times when processing social skills is extremely instructive, especially when students are working on extended team projects. But there are many cooperative interactions for which it would be very inefficient to define and process a social skill. The teacher without structures has a dilemma: cumbersome definition and processing of social skills for each activity or letting many activities slide, even though they do not satisfy the definition of cooperative learning in either the Learning Together or Kagan models. Teachers versed in Kagan Structures can move efficiently from activity to activity, assured they are creating solid cooperative learning at any moment. Teachers not versed in the Kagan Structures can either include many interactions which violate the principles of good cooperative learning (in both models) or take time off the academic task to continually define and process social skills.

Sustained Implementation

There is no empirical research on sustained implementation in either the Kagan or the Learning Together models. Thousands of teachers receive training in both models, but no controlled research asks how many of those teachers are implementing what they have learned one year, five years, or even ten years later.

The research which demonstrates positive effects of the methods does not address the question of sustained implementation. It does not matter how much research one amasses to support the positive effects of an educational innovation if teachers in the real world do not use that innovation or if they use it for only a few years and then drop it in favor of next year's new innovation. What good is a great method if it is abandoned after a few years?

Both experience with the Kagan Structures as well as theory of sustained implementation support the conclusion that the Kagan Structures are resistant to being dropped in the face of new educational innovations.

Experience with Kagan Structures

Individual teachers, schools, and districts have found the structures become a stable part of each teachers' repertoire and report sustained

implementation. When we return to very large districts in which we conducted training in the structures over ten years ago, a great many teachers are still using the structures and have taken it upon themselves to train new teachers so there is more work being done with the structures than at the end of our initial training.

Many examples of sustained implementation are posted on this web page. To review them, click here. To take but two of many examples,

We have been implementing Cooperative Learning in Davis District and the state of Utah now for over ten years. The impact and flow has rolled from one district to another as teachers learn new strategies for teaching.

–Colleen C. Uhl, Bountiful Elementary School, Bountiful, Utah

We still use, endorse, purchase, and spread the "structural words" at Lamplighter. Coleta and I have provided our teachers with all "The Smart Cards" and have had most satisfying results this year. The language of cooperative learning, the structural approach, resounds throughout our school! 2000 marks 10 years with Kagan for the Lamplighter School!

–Sheila Leventhal McCartor,The Lamplighter School, Dallas, Texas

Theory of Sustained Implementation

Central to the development of Kagan Structures was the desire to create methods which could be used to more efficiently deliver any new curriculum or educational innovation. Education has been plagued with a "Replacement Cycle" in which each new innovation in turn is replaced by next year's new thing. The replacement cycle is fueled by complex lesson-based innovations. If an innovation is so complex it cannot be implemented along with the next new educational innovation which comes along, it has a half-life. Why? Educational innovation is inevitable. The structures are not one more innovation, but a better way to implement any innovation. The structures are in contrast to complex lesson designs which compete for teacher time and energy.

For example, let's say a teacher has learned how to implement complex cooperative learning lessons. These lessons take time and energy to design. After a few years, along comes some new innovation such as multiple intelligence, brain-compatible instruction, emotional intelligence, standards-based instruction, higher level thinking, or technological literacy. If the new innovation requires complex lessons, complex cooperative learning lessons get dropped in favor of the new lessons. The teacher simply does not have time to design and implement complex lessons of both types.

If instead, however, cooperative learning is based on simple structures which can be used at the drop of a hat with no special planning or preparation, the teacher does not abandon the structures in the face of a new innovation. Rather the teacher uses the structures to more efficiently implement the new innovation. I am fond of saying that the Kagan Structures are not one more program with a half-life, destined to be replaced by next year's new innovation; they are a better way to deliver any program.

In Sum

Both the LT and Kagan models are powerful and transform learner outcomes in many favorable ways. They have considerable overlap. They do, however, differ in some important ways and implementation of one takes a quite different form than implementation of the other. Which model a teacher, school, or district adopts has important consequences – both immediate and long term. It is my hope that those deciding how best to instruct the nation's youth will read deeply in many theories, consider the needs of their teachers and students, and make an informed decision.