Groups Solve Problems Better Than Individuals?

(Un)interesting new research says that groups solve problems better than individuals.

Groups of three, four, or five perform better on complex problem solving than the best of an equivalent number of individuals, says a new study appearing in the April issue of the Journal of Personality and Social Psychology, published by the American Psychological Association (APA). This finding may transfer to scientific research teams and classroom problem solving and offer new ways for students to study and improve academic performance, according to the study authors.

In this study 760 students from the University of Illinois at Urbana- Champaign solved two letters-to-numbers coding problems as individuals or as groups of two, three, four and five people. Previous research has shown that groups perform better than the average individual on a wide range of problems. However, this study tested the relationship between group size and performance as compared to that of an equivalent number of individuals by comparing the number of trials to solutions and answers given for complex problems.

So in other words, they performed a tasks that most people don't perform on a regular basis. Hence then compared groups to individuals when both parties were equally ignorant of the solution. I bet that individual cryptographers would outperform these groups.

How to Cut the Legal Bill for a Small Business (Without Cutting Any Corners)

Maybe it is time for a post that details the circumstances under which groups are successful, and those under which they aren't. Sounds like a lot of work though…

Hat Tip to Geekland.

  • kbzimm

    See: The Wisdom of Crowds, published in 2004.

  • Rob

    Either you didn’t read the book, or you didn’t read the research I linked to, because they are incompatible. The “wisdom of crowds” only works when crowd members act independent of each other. When they work together they usually fail. A better model would be Robert Wright’s “Nonzero” which says that humans work well in groups when they plan nonzero sum games and all parties benefit.

  • Its better to work in groups as many ideas will flood in. The more the better. But u cant underestimate the power of individual mind. Ya it might be tricky to use groups for solving problems. Depending upon the problem.

  • I feel that for some people working in a group can be a disaster. I love indivualism and freedom associated with it. Most of the, when I have worked in a group, I have ran into problem. So, for me it does not work out well. I just feel that working in a group takes away much of my energy.

  • DavidG

    Strong parallel to your earlier post about experts v. crowds. Who is best at what might not be so hard to tease out.
    Try this hypothesis …
    1) Need expertise? Use an expert.
    2) Need a decision? Use a small team.
    3) Need an opinion? Use a crowd.
    1) expertise is the single BEST way of solving a problem,
    2) decision is the optimal of multiple potential solutions and,
    3) opinion is an unproven solution, most likely to succeed.

  • Rob

    I think that is a great start. The key is understanding what kind of problem you have.

  • If only it were anything close to that simple.

    Who’s an expert? How do you select one? How do you use a small group and avoid group think (or the “counting horse” bias)? And crowds… how do you seperate the wheat from the chaff?

    As to this research… they used problems where there is “one” right answer. That’s not the case in business. Nor is failing to have the right answer the sole reason that an answer fails.

  • It is understandable why groups are better than single minds. They discuss, they put together pieces of information and their ideas, they reach to the best decision by providing answers to different questions, while a single person can be wrong even if he thinks his solution is the best one.