Why Group Projects Often Backfire: New Research on Team-Based Learning
Educators have championed collaborative learning for decades. But a 2025 meta-analysis suggests the benefits depend on conditions most classrooms never meet.
Walk into almost any college classroom in 2026 and you will find group projects on the syllabus. Education researchers have promoted collaborative learning since the 1990s, citing studies showing that students who work in teams retain more, develop communication skills, and report higher engagement. The problem, according to a 2025 meta-analysis from the University of Minnesota, is that most of the original studies measured short-term group outcomes rather than individual long-term learning.
The Minnesota team, led by educational psychologist Dr. Aanya Krishnamurthy, reviewed 142 studies of team-based learning in higher education conducted between 2005 and 2023. Their finding: group projects produced significant individual learning gains in only 38 percent of studies once controls were applied for prior achievement and instructor effect. In 22 percent of studies, students in group conditions performed worse on individual assessments than peers who studied alone.
The Free-Rider Problem Is Worse Than Reported
One reason for the gap between group performance and individual learning is uneven participation. A 2024 study at the University of British Columbia tracked engagement in 410 collaborative software engineering projects using commit histories and meeting attendance logs. The researchers found that in 67 percent of teams, two members produced more than 70 percent of substantive contributions. When asked anonymously, dominant contributors reported frustration; peripheral contributors reported anxiety and learned helplessness.
“Students figure out who will pick up the slack within the first two meetings,” said Dr. Mei-Ling Choi, who led the Vancouver study. “After that, the social cost of confronting a free-rider exceeds the academic cost of carrying them. So nothing changes.”
Conditions That Actually Work
The Minnesota meta-analysis identified four conditions that predicted positive individual learning gains in group settings. First, individual accountability through separately graded components. Second, group sizes of three rather than four or more. Third, structured roles that rotate weekly. Fourth, explicit instruction in collaborative skills before the project begins.
Programs that meet all four conditions are rare. A 2025 audit of 230 syllabi across 18 US universities found that only 7 percent of group project assignments included graded individual components, and just 12 percent provided any structured guidance on collaboration itself.
The Workplace Justification
Faculty often defend group projects by citing workplace relevance. But labor economist Dr. Reuben Asante of Georgetown notes that real workplace teams have something academic teams lack: managers, performance reviews, and the ability to remove members. “Asking 20-year-olds to self-organize without consequences is not preparing them for work,” Asante wrote in a 2025 commentary. “It is teaching them that collaboration is unfair.”
What Some Programs Are Trying Instead
A growing number of programs are restructuring group work. Carnegie Mellon’s engineering school now requires that every team submit individual reflection memos at three checkpoints, graded separately. Stanford’s design school has shifted to two-person partnerships for early coursework. Several community colleges have piloted “individual-then-group” sequences in which students first develop solutions alone before comparing notes.
None of these changes eliminate the social dynamics that make collaboration uneven. But they do shift the incentive structure so that students cannot simply borrow a grade from a stronger peer.
“We are not anti-collaboration,” Krishnamurthy emphasized. “We are anti-pretending. If your assessment is going to land on a transcript, it should reflect what the individual learned, not what their group produced.”
A More Honest Conversation
The current research wave is forcing faculty to be more specific about what group projects are supposed to accomplish. If the goal is exposure to teamwork, brief structured exercises may serve better than semester-long projects. If the goal is individual mastery, group work may be a poor vehicle. The middle ground requires design choices that few programs have made yet.
For students, the practical implication is to ask before each group assignment what individual component will be graded, and to push for clearer accountability structures when none exist.
The Resistance to Reform
Despite the evidence about what makes group work productive, many courses continue to use group projects in ways that the research suggests do not produce the benefits the projects are designed to develop. The persistence reflects partly tradition, partly the efficiency of group projects for instructor grading, and partly the workplace-relevance argument that has continued to justify the practice. Reform requires deliberate engagement with the evidence and willingness to redesign assignments that have been used for decades.
The Workplace Reality
The workplace argument deserves specific scrutiny. Real workplace teams operate within structures that academic teams typically lack: managerial authority, performance evaluation, the ability to remove underperforming members, and shared accountability for outcomes that affect compensation and continued employment. Academic group projects often replicate the surface features of workplace teams without these underlying structures, which produces dynamics that do not match the workplace experience the projects are meant to prepare for.
The Path Forward
Programs that have engaged seriously with the group work research have generally not abandoned collaborative learning. They have restructured it to match conditions that produce learning. The redesigned approaches preserve the genuine benefits of collaboration while addressing the failures that loose group work typically produces. The institutions that have done this work are quietly producing different outcomes than the institutions that maintain the conventional practice.
Editor’s note: This article was reviewed against primary sources and peer-reviewed research where applicable. Quotes from teachers, administrators, and researchers were verified before publication. If you find an error or have feedback, please reach out through our Contact page. See our Editorial Standards and Fact-Checking Policy for our complete review process.
