What Makes Group Projects Work?

Few parts of university life cause as much complaint as the group project. Students grumble about classmates who do nothing, about meetings that go nowhere, and about marks that do not reflect their own effort. Yet employers consistently say they want graduates who can work in teams. Researchers have therefore been trying to work out what separates groups that succeed from those that fall apart.

Dr Helena Marsh of Corran University studied 120 student groups over two years. She found that the size of the group mattered more than almost anything else. Groups of three or four produced the best work; in groups of six or more, at least one member usually contributed very little. "In a large group, it is easy to hide," she says.

Professor Tomas Ridell takes a different view. In his research, the most important factor was not size but how roles were decided at the start. Groups that spent their first meeting agreeing who would do what, and writing it down, were far less likely to report conflict later. Ridell also found that groups in which students chose their own roles performed better than groups where the lecturer assigned them.

Dr Priya Castellan has focused on the problem of unequal effort. She tested a system in which students privately rated each other's contribution, with the ratings used to adjust individual marks. The system reduced complaints considerably, although Castellan notes that some students felt uncomfortable judging their friends. She also warns that peer ratings only work if students trust that they will remain confidential.

For Dr Ibrahim Solberg, the key issue is communication between meetings. By analysing the messages exchanged by project groups, he found that successful teams did not necessarily communicate more often, but they replied more quickly and asked each other more questions. Solberg believes universities should teach basic online communication skills, which many students wrongly assume they already have.

Finally, Professor Lena Okafor-Wren has examined mixed groups, in which students from different countries or academic backgrounds work together. Such groups often struggle in the first few weeks, she found, but by the end of a semester they frequently outperform groups whose members are all similar. The early difficulties, she argues, force members to explain their ideas more clearly.

Taken together, the research suggests that group projects are not doomed to fail. With sensible design, they can teach exactly the skills that employers say they value.