Group Projects That Actually Teach Something

Group work teaches well under two conditions, and fails under almost any others. The first is that students genuinely need each other to finish the task. The second is that each student is held responsible for learning the material personally, not just for the group's product. Decades of research on cooperative learning come back to this pair, and most bad group projects are missing one or both.

The researchers most associated with that finding are David and Roger Johnson, brothers at the University of Minnesota who have studied cooperative learning since the 1970s. They call the first condition positive interdependence and the second individual accountability. Their model adds three supporting elements: students talking face to face and helping each other, being taught the social skills that cooperation needs, and the group reviewing afterward how well it worked together. Their reviews, and others by researchers such as Robert Slavin at Johns Hopkins, have generally found that well-structured cooperative learning produces higher achievement than students working alone or competing against each other. The word doing the work in that sentence is "well-structured". A poster assigned to four students with a single shared grade is not what the studies tested.

Methods with a track record

The best known is the jigsaw classroom, designed by social psychologist Elliot Aronson in Austin, Texas, in 1971. Austin's schools had just been desegregated, and classrooms were tense and competitive. Aronson's team split each lesson into pieces. Every student in a small group got one piece, say one period of a historical figure's life, and met with students from other groups who had the same piece to master it. Then each returned to the home group to teach it. No one could pass the end-of-lesson quiz without listening to every member, so the structure itself created interdependence. Aronson and colleagues reported that students in jigsaw classes came to like their classmates more and that minority students' performance improved, though later replications have produced more mixed results on achievement.

Think-pair-share is smaller and quicker, and it is usually credited to Frank Lyman at the University of Maryland around 1981. The teacher poses a question, students think silently for a set time, discuss their answer with a partner, and then some pairs share with the class. The silent thinking period is the part that matters. It gives slower or quieter students time to form an answer before the fastest voice in the room fills the silence, and it means every student has to commit to an idea, even privately.

Assigned roles are the third common tool: a facilitator who keeps the discussion moving, a recorder, a skeptic or checker who asks whether the group can justify its answer, a timekeeper. Roles help most when they are tied to the thinking rather than just the logistics, and when they rotate. A student who is always the recorder is mostly practicing handwriting. Slavin's Student Teams-Achievement Divisions (STAD) adds another lever: teams earn recognition based on how much each member improves on individual quizzes, so the strongest student gains by making sure the weakest one learns.

The coasting problem, and grading individuals in a group

Anyone who has done group work in school or at a job knows the pattern. One person does most of it, one does a little, and one coasts. Psychologists call the general tendency social loafing; teachers often describe the student who does it as a hitchhiker or passenger. It is the most predictable way group projects go wrong, and it is mostly a design failure rather than a character flaw. If everyone receives the same grade and contributions are invisible, coasting is a rational choice, and the most conscientious student absorbs the cost.

Teachers who handle this well make each person's contribution visible and graded. Common approaches include a separate individual component, such as a quiz, a written reflection or a short oral explanation, that counts toward each student's mark. Others assign named sections with named authors, keep a shared document whose version history shows who wrote what, or call on a random group member to explain any part of the project, so everyone has to understand the whole. Peer evaluation is widely used too: students rate each other's contribution confidentially, and the teacher adjusts individual grades. It works better when students rate specific behaviors, such as meeting deadlines or contributing ideas, than when they give a single overall score, and when the teacher treats the ratings as one input rather than the final word.

Group size matters as well. Pairs and groups of three or four leave less room to hide than groups of six, and many cooperative learning guides recommend keeping groups small for exactly that reason.

Questions a parent can ask the teacher

If your child keeps coming home frustrated by group projects, or keeps doing everyone else's share, a calm conversation with the teacher is more useful than a complaint about a single grade. The aim is to understand how the project is built. These questions tend to get specific answers:

• How is my child's individual learning assessed, separately from the group product?
• How were the groups formed, and do they change between projects?
• Are roles assigned, and do they rotate?
• Is there a peer evaluation, and how much does it affect the grade?
• What should a student do if a group member is not contributing, and when should they tell you?

Most teachers have thought about these issues and will explain their system. If the answers suggest everyone simply gets one shared grade, it is reasonable to ask whether an individual component could be added next time.

At home, the most practical help is to ask your child to explain the parts of the project that someone else did. If they can do that, the group work is teaching them something. If they can only describe their own slice, that is useful information to bring to the next parent-teacher conference, along with a copy of the assignment sheet.