How to Teach Collaboration Skills in the Classroom

By the third week of school you can name them without looking at a seating chart. The child who builds the whole thing himself because waiting for anyone else is unbearable. The one who hands over the marker and drifts toward the window. The pair who spend eleven minutes negotiating who gets to write the title.

We seat four students at a table, call it group work, and hope something social happens. Then we grade the poster.

Collaboration skills in the classroom are not a personality trait that some children arrive with in September. They are a set of teachable behaviors, and they respond to explicit instruction the same way letter formation and scissor grip do. What follows is the progression I use with PreK through fourth grade, including how to keep it working for the students on an OT or special education caseload.

Why collaboration skills in the classroom need to be taught, not assumed

The strongest evidence here is old and unglamorous, which is usually a good sign. Cary Roseth, David Johnson, and Roger Johnson pooled 148 studies covering more than 17,000 students across eight decades of research. Students working under cooperative goal structures, where the group succeeds together or not at all, came out ahead of students in competitive and individualistic structures on achievement and on peer relationships at the same time. Most of that sample is early adolescents, so read it as a direction rather than a precise number for kindergarten. The direction is remarkably stable.

The social side may matter more than the academic side. Mark Van Ryzin and Cary Roseth ran a randomized trial across 15 middle schools with about 1,460 seventh graders. Classrooms using structured cooperative learning reported less bullying, less victimization, and lower stress among marginalized students, plus fewer emotional problems and a stronger sense of belonging across the whole group. That is a classroom management result, not a warm feeling.

Joseph Durlak and colleagues found the same pattern at scale. Across 213 school programs reaching more than 270,000 students, teaching social and emotional skills directly produced an academic gain equivalent to 11 percentile points. Skills that get taught get learned, and then they show up on the test.

The catch sits inside the word structured. Robyn Gillies, reviewing decades of cooperative learning research, is blunt: putting children near each other does not produce cooperation. Groups improve when the task genuinely requires everyone, when the social skills are taught instead of assumed, and when the group stops afterward to examine how it worked. The Pedagogy of Play team at Harvard Project Zero adds a related observation about playful learning, that children practice balancing independence and interdependence when the work feels like play rather than compliance.

Five steps to teach collaboration skills in the classroom

Run these in order. Most of the trouble I see in classrooms comes from starting at step three.

Step 1. Start with partners, not teams

Two students is a conversation. Four is a committee, and committees need procedures nobody has taught yet. I spend the first two weeks of any collaboration push entirely in pairs, then move to threes, then fours.

Materials: whatever you already have. Two students, one shared object.

• Setup: 1 minute. Partner them yourself. Do not let students choose yet.
• Time: 10 to 15 minutes, twice a week.
• What mastery looks like: both students touch the materials, and neither one narrates the entire activity.

Step 2. Pick a task one child physically cannot finish alone

This is the step people skip, and it is the one that does the actual work. If a single motivated student can complete the task, a single motivated student will, and everyone else learns that group work means watching.

Large scale building solves this by physics rather than by rules. A dome, a tunnel, or a doorway taller than the builders requires more hands than one child has. Life sized panel systems like Superspace work well for this because the pieces are big enough that a single pair of arms simply runs out of reach partway through the wall. Wide butcher paper murals, floor sized maps, and long block runs do the same thing on a smaller budget.

• Setup: 5 minutes to clear floor space. You need more room than you think.
• Time: 20 to 30 minutes for a first build.
• What mastery looks like: at least one student asks another student for help without looking at you first.

Step 3. Name the roles out loud before anyone touches materials

Unnamed roles get claimed by whoever is loudest. Named roles get rotated. I use four: builder, checker, supply runner, and reporter. Say them, assign them, and post them where students can see.

Rotate roles every session, not every unit. Students need enough repetitions in each role to get past the novelty, and the child who is always the reporter never learns to build.

• Setup: 2 minutes of assigning and repeating back.
• Time: adds about 3 minutes to any activity.
• What mastery looks like: a student corrects a peer by role ("you're the checker, is it straight?") rather than by name calling.

Step 4. Teach the exact words for disagreeing

Most group conflict in early elementary is not a values clash. It is a vocabulary gap. Children know how to say no and they know how to give in, and they have nothing in between.

Give them three sentence frames, model each one badly and then well, and practice them in a low stakes moment before you need them. I use: "I have a different idea, can I show you?" and "Can we try both?" and "I don't like that, but I'll try it your way first."

• Materials: three frames on a chart or a card taped to the table.
• Time: 10 minutes to introduce, 30 seconds to review before each build.
• What mastery looks like: you hear a frame used by a student when you are across the room and not watching.

Step 5. Spend the last two minutes on how it went, not what got built

Gillies calls this group processing, and it is the piece that gets cut when the schedule slips. Do not cut it. The reflection is where the behavior becomes a skill instead of an accident.

Two questions, asked to the group and not to individuals: what did your team do well today, and what will you do differently next time. Keep it to two minutes. Longer turns into a class meeting, and the energy drains.

• Time: 2 minutes, every single session.
• What mastery looks like: students name a behavior ("we let Malik finish") instead of an outcome ("we built it fast").

If you want tasks that already pass the step two test, I collected a dozen of them in a companion post: group activities that build collaboration skills in the classroom, with materials, setup times, and adaptations for each one.

How to adapt collaboration work for mixed needs

This is the part that decides whether collaboration is real or performative. If the activity only works for students who can stand, speak, and tolerate noise, it is not a collaboration lesson. It is a sorting mechanism.

The sensory seeker and the sensory avoider in the same group

Large builds tend to be loud and physical, which is exactly why seekers thrive there. Proprioception is the body's sense of where its own limbs are, and it is fed by pushing, pulling, and carrying weight. Give the seeker the supply runner role and let the heavy work regulate them.

For the avoider, protect the perimeter. Assign a role that happens at the edge of the build: checker, blueprint holder, or the student who approves each section before it goes up. Same task, same team, different sensory load. Noise reducing headphones stay available without a conversation about them.

Students who use wheelchairs, walkers, or gait trainers

Build horizontally on the floor or vertically at seated height, not at standing height. If the structure goes up, the seated student ends up handing pieces to people who are doing the interesting part.

The fix is boring and effective: set the build zone at the height of the seated student and make everyone else come down to it. Kneeling is fine for the rest of the group. Also plan the approach path before students start, because a build that fills the floor becomes a wall.

Nonspeaking students and emerging communicators

Train the peers, not just the student. Connie Kasari and colleagues ran a randomized trial in general education classrooms and found that a peer mediated approach, where classmates were coached in how to include a child with autism, produced better social network standing, more friendship nominations, and less playground isolation than working with the child alone. Those gains held at the three month follow up.

Practically, that means teaching the group how to wait, how to offer a choice of two, and how to read a device or a card. Load the student's AAC device or communication board with the words the build actually needs: more, done, turn, my idea, help, stop.

The student who takes over

Scarcity does what lectures cannot. Give each student a distinct portion of the materials and require that every portion appear in the finished structure. The dominant builder now has to negotiate for pieces instead of collecting them.

If that is not enough, make them the reporter for three sessions running. Explaining someone else's design decision is a fast education in listening.

Frequently asked questions

How long should a collaboration lesson take?

Fifteen to twenty minutes for PreK and kindergarten, twenty to forty for grades one through four, plus the two minute reflection. Frequency beats duration. Three short sessions a week will move a class further than one long block on Friday.

What age can students start working in groups?

Pairs work by age four with a task that has an obvious shared goal. True group negotiation, where children divide labor and resolve disagreements without adult refereeing, generally shows up around six or seven. Before that, expect parallel work with moments of genuine cooperation, and treat those moments as the win.

How do you grade collaboration?

Grade the process on a simple rubric and grade the product separately. I score four things: contributed ideas, listened without interrupting, shared materials, and stayed with the group through a problem. Three point scale. Two minutes per group while they are cleaning up.

What do you do when one student refuses to join the group?

Shrink the group to one partner and shorten the task. Refusal is almost always about volume, unpredictability, or a history of being talked over, and all three get better with a smaller container. Give the student a specific job with a clear endpoint rather than an open invitation to participate.

Do collaborative building activities work for students with IEPs?

Yes, and the research on cooperative structures suggests the students furthest from the social center often gain the most, as Van Ryzin and Roseth found with marginalized students in their trial. The requirement is structure: assigned roles, a task that needs every person, and a reflection at the end. Unstructured group work tends to widen the gap rather than close it.

Where to start on Monday

Pick one pair activity, run it twice this week, and end both sessions with the same two questions. That is the whole thing. Add a role, then a build too big for one child, then a sentence frame, and by October you will have a class that argues productively instead of quietly quitting.

References

Durlak, J. A., Weissberg, R. P., Dymnicki, A. B., Taylor, R. D., & Schellinger, K. B. (2011). The impact of enhancing students' social and emotional learning: A meta-analysis of school-based universal interventions. Child Development, 82(1), 405 to 432. https://doi.org/10.1111/j.1467-8624.2010.01564.x

Gillies, R. M. (2016). Cooperative learning: Review of research and practice. Australian Journal of Teacher Education, 41(3), 39 to 54. https://doi.org/10.14221/ajte.2016v41n3.3

Kasari, C., Rotheram-Fuller, E., Locke, J., & Gulsrud, A. (2012). Making the connection: Randomized controlled trial of social skills at school for children with autism spectrum disorders. Journal of Child Psychology and Psychiatry, 53(4), 431 to 439. https://doi.org/10.1111/j.1469-7610.2011.02493.x

Mardell, B., Wilson, D., Ryan, J., Ertel, K., Krechevsky, M., & Baker, M. (2016). Toward a pedagogy of play: A Project Zero working paper. Harvard Graduate School of Education. https://www.pz.harvard.edu/resources/toward-a-pedagogy-of-play

Roseth, C. J., Johnson, D. W., & Johnson, R. T. (2008). Promoting early adolescents' achievement and peer relationships: The effects of cooperative, competitive, and individualistic goal structures. Psychological Bulletin, 134(2), 223 to 246. https://doi.org/10.1037/0033-2909.134.2.223

Van Ryzin, M. J., & Roseth, C. J. (2018). Cooperative learning in middle school: A means to improve peer relations and reduce victimization, bullying, and related outcomes. Journal of Educational Psychology, 110(8), 1192 to 1201. https://doi.org/10.1037/edu0000265