Are Two Heads Better Than One?
It depends on the heads and how they interact.
Posted Feb 14, 2016
According to the adage, “Two heads are better than one.” But is it true that we can work more intelligently as a team than as individuals? New research by Carnegie Mellon University professor Anita Williams Woolley and her colleagues suggests that the intelligence of a group can exceed that of its members if the right conditions are met.
First we need to get a handle on what psychologists mean by “intelligence.” A century ago, British psychologist and statistician Charles Spearman found that individuals tend to perform similarly across a wide range of cognitive tests, even though they may do better on some tasks than others.
Let’s say we subject Candace and Caleb to a battery of verbal and quantitative tests. We find that Candace does better on the verbal than quantitative tests and the reverse for Caleb. However, Candace still does better than Caleb at both kinds of tests. According to Spearman’s theory, Candace has greater general intelligence than Caleb, even though their specific abilities differ. General intelligence is like the tide that raises both the yacht and the dinghy.
Intelligence includes more than just knowledge and skills. It also encompasses the ability to reason and to solve problems as well as to learn from experience. Speed of cognitive processing is an important component too. So intelligent people aren’t just book smart—they can also think quickly and effectively.
Traditionally, many psychologists have assumed that the intelligence of a group was no more than the average intelligence of the individual members. In other words, two heads may be able to do more work than one head, but two heads can’t work any smarter than either one could alone. Alternatively, some psychologists have suggested that the intelligence of the group is merely that of its most intelligent member, who dominates the group and coordinates its activity.
However, Woolley and colleagues’ findings show than neither of these beliefs is correct. In their lab, they had groups complete a set of cognitive tasks, much like the test batteries used to measure individual intelligence. They also measured the general intelligence of each individual group member.
Woolley and her associates discovered that the intelligence of the individual members didn’t predict how well the group would perform. Instead, they found that any particular group tended to perform similarly across tasks. And, of course, some groups tended to outperform other groups regardless of the test.
In other words, the group as a whole exhibited a kind of general intelligence. Woolley calls this “collective intelligence” to distinguish it from the general intelligence of individuals. These findings are in line with idea of extended cognition, which states that mental processes aren’t encapsulated inside the heads of individuals but extends to others working as teams. Thus, the team as a whole develops a group mind.
Woolley’s team has uncovered a number of factors that determine how well a group will perform. These findings will certainly be of interest to anyone who has to work collaboratively with other people. This pretty much includes all of us at one time or another.
The most important factor in determining how intelligently a group will perform has nothing to do accumulated knowledge or skill set. Rather, it depends on how well the individual members can read the emotions of the other group members. Woolley calls this “social perceptiveness.”
In one study, Woolley and colleagues first tested participants’ ability to judge facial expressions of emotions in photographs. The participants were then assigned to groups, which performed a battery of cognitive tasks. The average social perceptiveness of the individual members was the strongest predictor of how intelligently the group as a whole would perform. This was true even when the group members could only communicate through text messaging so that all of the usual indicators of emotion, such a facial expressions, posture, and intonation of voice, were missing.
While the average intelligence of the members does contribute to the group’s collective intelligence, social perceptiveness is more important. Thus, a group composed of members who have moderate intelligence but are very good at reading each other’s emotions can outperform a group with high average intelligence but low social perceptiveness.
Another important factor in collective intelligence is the degree of diversity, especially in terms of cognitive style or personality. Having members who approach tasks differently and have different skills and experiences can enhance the collective intelligence of the group—to a degree.
Too much similarity among members limits the range of approaches the group can apply to a task. Likewise, too much difference among members can lead to communication breakdowns and a lack of empathy for others in the group. Instead, a “Goldilocks” zone of not too much and not too little diversity leads to the best performance.
Collective intelligence is an example of what’s known as an emergent property, a characteristic of a system that doesn’t occur in any of its parts. Emergence isn’t solely a social-psychological phenomenon. Rather, it occurs regularly throughout the natural world.
A good example of emergence is water. Although water is a liquid, it’s composed of two gases, hydrogen and oxygen. Thus, water as a whole has characteristics—wetness, waviness—that neither of its components has. This is what is meant by an “emergent property.”
As psychology moves into the twenty-first century, we’re finding more and more examples of emergence in social interactions. We aren’t just, in philosopher Alan Watts’ words, “skin-encapsulated egos.” We are indeed individuals, but we’re also components within larger social systems that take on identities of their own.
Woolley, A. W., Aggarwal, I., & Malone, T. W. (2015). Collective intelligence and group performance. Current Directions in Psychological Science, 24, 420-424.
David Ludden is the author of The Psychology of Language: An Integrated Approach (SAGE Publications).