Having the smartest guys in the room does not guarantee a great brainstorm session or success at a problem-solving meeting. Instead, groups do best when they include people who take turns talking and have high social sensitivity — which means they tend to include women — according to a new study.
The findings could be used to revolutionize how governments, corporations and schools decide to organize the most effective groups to cooperate on challenges ranging from the complex impacts of climate change to homework, the researchers contend.
The study also gives the first solid measure of collective intelligence, dubbed the "c factor" by researchers and defined as a group's ability to perform a wide variety of tasks.
"All of this really calls into question our notion of intelligence; it is becoming less relevant what an individual can do by him or herself and more relevant to understand what individuals can accomplish through others and with technology," said study researcher Anita Woolley, an organizational behavior scientist at Carnegie Mellon University in Pittsburgh.
Woolley and her colleagues carried out two studies with 699 volunteers. The participants broke down into smaller groups of two to five individuals as they tackled a variety of cognitive tasks, such as solving puzzles, brainstorming, casting collective moral judgments, and negotiating over limited resources.
Simpler tests allowed the researchers to define certain domains of group performance and create a measure of collective intelligence. They then tested how well collective intelligence predicted a group's performance on more complex tests, including a checkers video game and an architectural design task.
Having higher average individual intelligence and the smartest individuals did have some connection to which groups would perform better. But the social factors proved a far better predictor of group performance.
Groups that included a higher proportion of women may perform better in part because women tend to demonstrate better social sensitivity, or the ability to read others' minds and moods, than men, according to the researchers.
Some surprises emerged when factors expected to make a difference for group performance did not deliver. For instance, group cohesion, motivation and satisfaction did not predict how well a group would perform.
Now researchers can begin applying their measure of collective intelligence to see how outside factors influence a group's performance, the scientists say. Organizations might finally get a handle on how Internet tools ranging from Facebook to instant messaging affect group performance – both in online and offline social networks.
"If we can say that collective intelligence is enhanced by X amount through the addition of certain tools, then we can better understand their contribution," Woolley told LiveScience in an e-mail.
Similarly, organizations that must constantly disband and form new groups can begin to understand the value that they lose (or not) in making such decisions. That may matter more than ever in today's world, where solving problems requires groups of individuals with deep expertise in different areas.
"The better our tools are for understanding how to effectively construct such groups, the better we will be at responding to those situations, whether they be in corporations, governments, schools, or the broader environment," Woolley said.
The study is detailed in the Sept. 30 issue of the journal Science.
