Being one smart cookie doesn’t matter much if you’re working in a group, according to the first study to calculate collective intelligence—a group’s ability to succeed at a variety of tasks.Surprisingly, in a team an individual’s smarts has little to do with success in thought-based tasks such as visual puzzles and negotiating over scarce resources, a battery of recent experiments found.
Instead, a group is more successful if it contains people who are more “socially sensitive”—in this case meaning they’re better able to discern emotions from people’s faces.
That also explains why groups with more women—who consistently score higher on tests of social sensitivity—were more likely to excel, said study leader Anita Williams Woolley, an expert in collective intelligence at Carnegie Mellon University in Pittsburgh.
Particularly intelligent groups also had more people who took turns speaking, according to the study, published tomorrow in the journal Science.
“There’s such a focus on individual intelligence and individual accomplishment, especially in western culture,” she said.
“As our world becomes flatter and more interconnected, it’s not as important to consider what an individual can do by themselves but what they can do collectively.”
Calculating Collective Intelligence
Individual intelligence is measured by the ability of a person to do multiple tasks well, Woolley said. To see if the same would be true of groups, Woolley and colleagues recruited 699 volunteers and first measured each person’s intelligence and social sensitivity using standard psychological tests.
The volunteers were then randomly split into groups of two to five and asked to do some simple tasks, such as solving a visual puzzle.
The results from the first round of experiments revealed that certain groups were better at all types of tasks, which is the “primary evidence for the notion of collective intelligence,” Woolley said.
By tallying the scores on this first barrage of tests, the scientists came up with a collective intelligence score for each group.
Next, the groups were each asked to perform more complex tasks, which included playing a video game against an imaginary opponent and solving a research-and-development problem.
As suspected, the groups’ collective intelligence scores from the first round of tests predicted how they’d do on the complex experiments, Woolley said.
Findings May Benefit Real-World Groups
This ability to predict group success may offer guidance in real-life situations—especially as more decisions in fields such as business and the military are made in consensus-based settings, she said.
For instance, knowing a group’s collective intelligence could be crucial in a high-risk situation where “suboptimal performance would be costly,” such as embarking on a new business venture, she said.
Likewise, the findings could help people succeed in group therapy, according to Bonnie Jacobson, a clinical psychologist in New York City.
Already the study’s conclusions confirm the group-based problem-solving that Jacobson has observed while leading therapy sessions.
“There’s no individual therapist that ever lived that has the brilliance of a group—if it’s Freud, Jung, Erikson,” she said.
“That’s what they’re proving here—I just love it.”
Group Smarts Rooted in Animal Kingdom
Group savvy may also be rooted in evolution, study leader Woolley said.
For instance, “you can’t be really good at hunting but not good at gathering,” she said.
“Figuring out how to more flexibly deploy the skills available through members of groups would be associated with survival.”
(Related: “Future Humans: Four Ways We May, or May Not, Evolve.”)
Likewise, Woolley said, the study is evidence that human societies may function better en masse, just like many animal species.
“It’s no mistake that some of the earlier work on collective intelligence does borrow from the animal kingdom,” she said.
(Read more about animal minds in National Geographic magazine.)
For instance, she said, “ants are simple creatures but collectively can accomplish things that are amazing.”