Individuals may outperform groups when it comes to brainstorming for ideas (see earlier post), but for logic-based problem solving, it seems three-person groups work best.
That’s according to Patrick Laughlin and colleagues who tested 760 students on a series of letters-to-numbers problems. Such problems involve the numbers 1 to 10 being allocated to the letters A to J, and the task is to find out which letters refer to which numbers in as few trials as possible. On each trial, equations (e.g. A + B = ?) are put to the researchers who will provide the answer in letter form, and the students can then make a guess for what that letter stands for, with the researchers saying whether the guess is true or false.
For each of 40 letter-to-number problems, the performance of students working alone was compared with the performance of a two-, three-, four- and five-person group working on the same problem.
The two-person groups didn’t tend to perform any better than the best of two students who were working alone on the same problem. However, the three, four and five-person groups consistently outperformed the best of three, four or five individuals working alone on the same task as them. The groups solved the problem more quickly and used more sophisticated equations.
However, the four- and five-person groups were no better than the three-person groups, suggesting a team of three is the optimum group size for logic-based problem solving.
“If groups of three perform as well as groups of larger size, it is obviously a more efficient use of human and logistic resources to use three-person groups”, the researchers concluded. “Further research should be conducted to determine whether three persons are necessary and sufficient for groups to perform better than the best of an equivalent number of individuals on other problem solving tasks”.
Laughlin, P.R., Hatch, E.C., Silver, J.S. & Boh, L. (2006). Groups perform better than the best individuals on letters-to-numbers problems: Effects of group size. Journal of Personality and Social Psychology, 90, 644-651.