Kagan Online Magazine

Subscription Info

Subscribe Me
You will receive an e-mail with hot links to the latest Kagan Online Magazine, as well as announcements about upcoming Kagan professional development events and new products.

Subscribe A Friend
Please include your friend's e-mail address when subscribing.

Unsubscribe Me
Remove yourself from Kagan's e-mail list.

Bookmark and Share

Spencer’s Thinkpad

Pairs vs. Teams

The Necker Cube of Cooperative Learning

Dr. Spencer Kagan

To cite this article: Kagan, S. Pairs vs. Teams: The Necker Cube of Cooperative Learning. Kagan Online Magazine, Issue #63. San Clemente, CA: Kagan Publishing. www.KaganOnline.com

Text Box: The Necker CubeTeachers sometimes ask me, When is it better to have students in pairs vs. teams? The answer is a bit like the Necker Cube: look at the cube one way, and the light purple face appears in front. Look at the cube another way, and the black face appears in front. The same is true for pairs vs. teams. Look for a while at the advantages of pairs, and pairs look better than teams; but look at the advantages of teams, and teams look better than pairs!

This article addresses five topics:
1. Why teams, not pairs, are the usual way to structure cooperative learning.
2. The advantages of teamwork.
3. The advantages of pair work.
4. The steps of five powerful pair structures.
5. A suggestion regarding pairs vs. teams based on the needs of today’s youth.

Cooperative Learning Has Emphasized Teams

In the history of cooperative learning, there was initial heavy emphasis on teamwork. Pair work was seldom considered. Most of the research that finds cooperative learning outperforms whole-class instruction is based on teams. Cooperative learning became identified with the use of teams, not pairs. One approach to cooperative learning was called Student Team Learning. Probably for these reasons, teamwork rather than pair work became the norm.

Advantages of Teamwork

Having students in cooperative learning teams has many proven advantages over teaching to a whole class. The research-proven advantages include improved academic achievement, reduced racial and academic achievement gaps, improved social skills and relations, reduced discipline problems, and improved race relations. Cooperative learning in these studies has consisted mainly of teamwork, although a few studies include students working in pairs within teams. To my knowledge, no studies have directly compared the advantages of teamwork over work in pairs. We can, however, contrast teams and pairs using purely theoretical considerations. Among the advantages of teamwork over work in pairs are:

  • Broadened Perspective
  • Increased Creativity
  • Team Cohesion and Belonging
  • Enhanced Problem Solving
  • High Achieving Models
  • Management Advantages
  • Improved Authentic Assessment
  • Enhanced Stimulation and Excitement

Broadened Perspective. If we want students to consider something from many points of view or include many ideas, teamwork is better than pair work. Examples include Team Statements and Sum-the-Ranks. Having four points of view to consider before making a decision will result in a wiser decision than considering only two points of view. Teams create more diversity than pairs.

Creativity. Related to the broader perspective provided by four minds rather than two is the enhanced creativity that results from teamwork compared to pair work. Synergy occurs as students build off the ideas of each other. One idea sparks a new, related idea. In teams of four, there are more ideas generated, so creativity is enhanced.

Team Cohesion and Belonging. Another consideration is team cohesion and student belonging. Team structures build team cohesion and team identity in ways that pair structures do not. Student teams have their own name and often their own cheer. Students identify themselves as a member of a team far more than thinking of themselves as a member of a pair: I am part of the Brain Trust team. Team cohesion is a powerful force to forge positive student relations, creating feelings of safety and belonging.

Problem Solving. If we have students solving problems in pairs, it is possible neither partner will be able to solve the problem. The same problems solved in a team run a far lower risk of having no teammate able to solve them because we advocate heterogeneous ability teams. With heterogeneous ability teams, the eight highest achievers in the class are spread out, one per team, maximizing the probability that one student on the team will know how to solve the problem and can tutor their teammates. We can, however, mitigate this pair work drawback with a management rule: Pairs Consult. The rule is simple: If students are working in pairs and are stuck or disagree on a procedure, they can consult with another pair.

Role Models. With the eight highest achievers spread out one per team, each team has a positive achievement role model. When we break students into pairs, only half the pairs have one of the eight highest achievers, so fewer students are in frequent contact with the highest achievers in the class.

Management. Managing eight teams is easier in many ways than managing sixteen pairs. For example, if there are materials involved, having eight Material Monitors pick up materials for their team takes less time and creates less confusion than having sixteen students lining up to pick up materials for themselves and their partners.

Authentic Assessment. It is much easier for a teacher to visit and view the work in eight teams than sixteen pairs.

There is simply more stimulation and more energy generated in teamwork than in pair work.

Stimulation and Excitement. Teamwork creates more excitement than pair work. Contrast the pair and team versions of the same structure: Numbered Heads Together and Paired Heads Together. Each student in the team or pair has a number. The teacher poses a question and each student on their own writes their answer, and then the students share their answers with their teammates or partners. The teacher then calls on one student in the team or pair to display their best answer, often with a finger response or by writing on and displaying their AnswerBoard. Finally, students in their team or pair celebrate. Showing your answer to teammates is more exciting that showing it to just a partner, and team cheers are far more exciting than a congratulations from a partner. There is simply more stimulation and more energy generated in teamwork than in pair work. We will return to this issue in the last section of this article, showing why stimulation and excitement are especially important for today’s youth.

Advantages of Pair Work

Although there are many advantages of teamwork over pair work, pairs offer their own advantages over work in teams. Among the advantages of pair work are:

  • Increased Active Engagement
  • Decreased Mind-Wandering
  • Safety
  • Intimacy
  • Management
  • Partner Variety

Although the advantages of pair work are fewer than the advantages of teamwork, the increased active engagement during pair work is tremendously important. In fact, it is perhaps the most important single advantage of all the advantages of team and pair work, and in some situations outweighs all other advantages.

Increased Active Engagement. From the beginning of our work in Kagan Cooperative Learning, we advocated teams of four. Other schools of cooperative learning were far less adamant about teams of four, advocating teams of three or being content with teams of five. Our insistence on teams of four was primarily because a team of four breaks neatly into two pairs. For pair work, a team of three or a team of five leaves someone without a partner. In Kagan Cooperative Learning we often tell students to turn to their face partner or shoulder partner to engage in a pair structure.

Although the advantages of pair work are fewer than the advantages of teamwork, the increased active engagement during pair work is tremendously important.

Our heavy emphasis on frequently breaking the team of four into two pairs is based on one of the basic principles of cooperative learning: simultaneous interaction. The principle of simultaneous interaction is unique to Kagan Cooperative Learning. To check the extent to which the principle is implemented, we ask, Are many students overtly interacting at the same time? It is like taking a snapshot of the class and counting how many students in that moment are talking or writing or doing some observable active engagement. The more students overtly interacting simultaneously, the greater the active participation.

The principle of simultaneous interaction tells us that pair work usually doubles overt active participation. For example, during a RoundRobin, at any one moment one of four students in the team is talking. During a RallyRobin, at any one moment one of two students is talking. This has extreme implications. To give each student in the class a minute to verbalize their ideas takes four minutes if we use RoundRobin, but only two minutes using a RallyRobin. If our goal is to have students verbalizing their ideas, we reach our goal in half the time using RallyRobin vs. RoundRobin. Let’s imagine we use both RoundRobin and RallyRobin repeatedly in our classrooms and over time we have used an hour of class time using each of these structures. How have students spent their time? Students have been actively engaged producing language and ideas for 15 minutes using RoundRobin, but they are actively engaged for 30 minutes of the hour using RallyRobin.

Yet another way to see the greater engagement generated by pair rather than team structures is to contrast student “down time.” If students are given two minutes each to share their ideas, in a RoundRobin they must spend six minutes of down time in which they are passively listening to teammates. For those six minutes they are not producing or verbalizing ideas. In RallyRobin, to give each student two minutes to share their ideas, students spend only two minutes of down time. Using a pair structure, we accomplish the same goal in half the time, and we also dramatically cut down the probability of mind-wandering!

In Kagan Cooperative Learning, we have developed many team structures and many pair structures, but the principle is the same. Pair structures generally double overt active participation compared to team structures. I say “generally” because there are some team structures in which all students are engaged at once, like in Jot Thoughts and Team Mind Mapping. Also, during specific steps of some team structures, like Numbered Heads Together, all students are active simultaneously, writing or solving a problem.

In Kagan Cooperative Learning, we weigh simultaneous interaction very heavily. As Kagan’s motto states, It’s all about engagement! A great deal of the research-proven gains of cooperative learning compared to whole-class instruction are due to increased active engagement. In the whole class, when one student is called on by the teacher to answer a question, only about 3% of classmates are actively engaged at that moment while many students are mind-wandering. When we turn the chairs around and have students in teams, when one student in each team answers, 25% of classmates are actively engaged at any moment. Following this logic, let’s often have students turn to their shoulder or face partner to interact, so at any moment 50% of classmates are actively engaged.

Decreased Mind-Wandering. In a Timed Pair Share in which students are given two minutes each to share, students spend two minutes waiting their turn to speak. But in a Timed RoundRobin in which students are also given two minutes each to speak, students spend six minutes waiting their turn to speak. Six minutes of passively listening to teammates speak dramatically increases the probability of mind-wandering. Mind-wandering is a well-researched, pervasive, covert form of off-task behavior. Mind-wandering occurs between 15% and over 50% of a person’s time, depending on the task.1 Mind-wandering during lectures occurs up to 50% of the time and is related to lower performance on exams.2 Mind-wandering rarely occurs during overt active engagement. It is unlikely to have one’s mind on something else while generating and verbalizing one’s ideas.

Safety. Years ago, I was involved in a ten-year project to train teachers in cooperative learning to promote language acquisition. During that time, I was doing an observation in a class in which the teacher used a variety of team structures for almost the whole class period. I was tracking the lack of participation of one boy. During the teamwork, the boy was silent. I began wondering if the boy might be mute or if he had little or no English language skills. At the end of the lesson, the teacher had the students stand up and pair up and share with a partner what they had learned in the lesson. To my great surprise, the boy was animated and shared several things he had learned! Based on additional observations, I realized that this student was not comfortable talking to a team of three, but was comfortable talking to just one partner. Students feel safer sharing in a team rather than in front of the whole class, but they feel even safer sharing with one partner rather than talking to a team of three others.

In language acquisition literature, the term “affective filter” refers to the negative impact on cognition and language production created by lack of safety. The notion of an affective filter is supported by brain research. When we do not feel safe, the amygdalae fire, sending inhibitory messages to the prefrontal cortex inhibiting the production of thoughts and language. Sharing with a partner reduces threat compared to speaking to a group.

Intimacy. Sharing with one other person results in sharing more intimate ideas than speaking in a group. This is easily observable by contrasting what is shared during Three-Step Interview vs. Team Interview. During Team Interview, one student at a time is interviewed by teammates. Often, answering to a group leads to far less intimate sharing than sharing with just one other person. During the first two steps of Three-Step Interview, students do pair interviews. They are physically closer and are talking directly to one person who is giving them full, undivided attention. Thoughts and feelings are shared at a deeper level.

Containing Disruptive Behavior. In classrooms in which there are many disruptive students, pair work is easier to manage than teamwork because four disruptive students on one team can spell disaster. They are likely to get off task and to encourage each other in disruptive behavior. In pairs, disruptions are contained to two students and are easier to manage.

Partner Variety. Students love variety. Pair work allows a greater variety of partners. The teacher calls for students to turn to their face partner. Later the teacher calls for students to work with their shoulder partner. Further, pair work occurs during many classbuilding structures like Quiz-Quiz-Trade and Inside-Outside Circle. Using a variety of pair structures contrasts with having the students always work with the same other three during teamwork.

Five Powerful Pair Structures

As I said at the outset, the history of cooperative learning biased our thinking, giving a preference to teamwork rather than pair structures. In hopes of correcting that imbalance, especially for educators who are not using many pair structures, I share here some of my favorite pair structures and why I like them.

Pair Share. Pair Share is one of the simplest of all the cooperative learning structures. As such it is among the easiest for teachers to learn and implement and easiest for students to perform. Essentially one partner shares, and then the other partner shares. To equalize participation, the teacher tells students they can share either one, two, or at most three sentences each in response to the question the teacher poses.

Contrasting Content for Pair Share vs. Timed Pair Share

Use Pair Share for brief responses and Timed Pair Share for longer, more elaborate responses.

Pair Share: In one sentence, name something that makes you happy.

Timed Pair Share: For one minute, describe in detail an event that made you feel happy.

Pair Share can be enhanced by having students respond to their partners after their partner shares. Here are three types of responses the teacher can call for after each partner shares:

  • Copycat Response. Students are told to say to their partner what the teacher tells them to say. Example: I enjoy listening to you.
  • Complete the Sentence Response. Teacher provides a sentence starter for students to complete. Example: What I like about listening to you is…
  • Paraphrase Response. Teacher asks students to paraphrase their partner. Example: An idea you shared is…

Timed Pair Share. Pair Share is excellent for short responses. Participation is equalized by the number of sentences each student shares. For longer responses, we use a Timed Pair Share. Each student shares for a predetermined amount of time. When we want longer responses, to equalize participation we need to have timed responses because—otherwise—in some pairs one student might do most of, or even all, the talking. Timed Pair Share may be done with no partner responses or with any of the three response types described above.

Pair Interview. Pair Interview is brief. The only difference between a Pair Interview and a Pair Share is that the partner of the person sharing asks a question or suggests a topic for their partner to talk about. Pair Interviews are usually limited to two- or three-sentence responses. Typical question: What about your project makes you proud?

RallyRobin. In RallyRobin, students take turns stating things to create an oral list, usually in a word, phrase, or a sentence at most. The oral lists are on topics like rainforest animals; inert elements; prime numbers; adjectives to describe the main character; foods you like.

RallyCoach. RallyCoach is a pair problem-solving structure. It works with just about any set of problems. Instead of solving problems solo, partners take turns. First, Partner A solves a problem while Partner B watches and listens, coaching if necessary. Then partners switch roles for the next problem. RallyCoach creates a high degree of support, active engagement, and individual accountability.

A Suggestion: Pairs and Teams

Some teachers like pair structures more than team structures primarily because of the greater active engagement and the simplicity of implementing and managing pair structures. Other teachers weigh more heavily the synergy and team unity generated by teamwork. Ultimately, this issue is ripe for empirical testing. What would be the results in classrooms that used pair structures exclusively, team structures exclusively, or a mix of both? Results could be analyzed regarding academic outcomes; student liking for class and content; and student relations, including social skills and race relations.

Posing the question of pairs vs. teams, however, creates a false choice. There is a very strong rationale for all teachers to use a healthy mixture of both pairs and teams.

Clearly there are times when teamwork is preferable and other times when pair work is preferable. Although all teachers will have their own unique situations and personal preferences, given the advantages unique to both pairs and teams, it seems we would all do well to use both. Without sometimes using pairs, we cannot maximize active engagement; without sometimes using teamwork, we cannot release maximum synergy and the power of team cohesion and identity. In Kagan Cooperative Learning we have found the best way to structure for the use of both pairs and teams: creating academically heterogenous teams of four but frequently having teammates break into two pairs for pair work, working sometimes with their face partner and sometimes with their shoulder partner.

There is also an overriding rationale for using a healthy mix of pairs and teams, especially in high school. Students have become accustomed to and enjoy a high level of stimulation. At a time when students are inundated with fast moving stimulation (video games, virtual reality, TikTok, texting, and social media), traditional classrooms by comparison are boring.

Indiana University's High School Survey of Student Engagement (HSSSE) reached more than 81,000 students in 110 high schools across 26 states, predominantly in the Midwest. Two out of three high-school students say they are bored in class every single day.3

A 2013 Gallup poll of 500,000 students in grades five through 12 found that nearly eight in 10 elementary students were “engaged” with school, that is, attentive, inquisitive, and generally optimistic. By high school the number dropped to four in 10. A 2015 follow-up study found that less than a third of 11th-graders felt engaged. When Gallup asked teens in 2004 to select the top three words that describe how they feel in school from a list of 14 adjectives, “bored” was chosen most often by half the students. Only two percent said they were never bored.4

A variety of ways of interacting with peers, including a healthy mix of pair and team structures, creates a higher level of stimulation and engagement than always using the same ways of structuring interaction in the classroom. Increasing the level of stimulation and engagement in our classrooms is a great solution to the school boredom epidemic.


1. Smallwood, J. & Schooler, J.W. The restless mind. Psychological Bulletin, 2006, 132(6), 946-958.

2Lindquist, S.I. & McLean, J.P. Daydreaming and its correlates in an educational environment. Learning and Individual Differences, 2011, 2(2) 158-167.

3Bryner, J. Most students bored at school. Live Science, February 2, 2007. https://www.livescience.com/1308-students-bored-school.html

4Jason, Z. Bored out of their minds. Harvard Ed. Magazine. Winter 2017. https://www.gse.harvard.edu/news/ed/17/01/bored-out-their-minds