Ideas & Findings

A new study concludes what most people have known for years: When it comes to solving problems, two (or more) heads are better than one.

So say the University of Minnesota researchers who reviewed some 50 years' worth of studies on cooperative learning.An abundance of research does point to the general effectiveness of cooperative-learning techniques in the classroom, but researchers argue over its usefulness for high-level problem-solving. To resolve that question, Zhining Qin, David W. Johnson, and Roger T. Johnson re-analyzed 43 studies conducted between 1929 and 1993. For the purpose of their analysis, the researchers classified the kinds of problems used in the studies in four ways: well-defined problems or those for which the problem-solving strategy is clear; ill-defined problems; linguistic problems or those solved through writing or talking; and nonlinguistic problems, which were mostly math problems.

For all four problem types, their meta-analysis found, members of cooperative-learning teams outperformed the students who were solving problems on their own. But cooperative-learning strategies were more effective for solving nonlinguistic problems than they...