The art and science of building a collectively intelligent team
Several years ago we had hired an HR consulting firm to help us assess our senior managers. Through a series of competency mapping exercises they were going to tell us who could be tasked with greater responsibility, who would have to go and what skill gaps we might fill by training. They had devised a set of exercises among which was a group problem solving task. In the group task they each assigned themselves one participant to observe, and made notes on the person’s participation and contributions. I sat in as an observer in one of the group exercises, not tracking any one person in particular.
Among the group was a manager whose assessment so far had been on the brink. Communication skills were below average, logical and quantitative skills were below average. At the end of the group discussion his assessment was no better, minimal participation. He might have to go, the consultants recommended. But I didn’t think so.
It was really the first time I had sat quietly observing a team engaged in a problem solving exercise. The consultants had their agenda but I was curiously watching something else – the dynamics. The team started out with the most obvious approach, a few very articulately arguing about whether it would work or not, many offering reasonable suggestions. This gentleman remained silent. And then he offered up a thought. It was not well thought out and a bit out there in left field. But it was just reasonable enough to get the articulate, logical folks unstuck enough to explore a different possibility. He did this three times in the discussion, each time he moved the group to a different solution space. For the rest of the time he sat silent. In the end the group emerged with an excellent solution quite far afield from where they had begun. Had he not been there to throw the inswings or curve balls, I imagine, the result would have been completely different. The group could not have made those large leaps to consider completely new approaches. For his part, he could not have seen those possibilities to full fruition. The outcome was greater than the sum of the parts.
People have long suspected that there is a different sort of intelligence that emerges in the collective, and an increasing body of scientific evidence is accumulating to describe it. A few years ago a group from Carnegie Mellon and the MIT Sloan Collective Intelligence Group published a study in the journal Science where they set out to see if they could create a collective intelligence quotient for the team, and more importantly, whether this was related to the average and maximum IQ of members of the teams. Based on the results of a series of problems given to almost 200 groups of 2 to 5 people, they derived a collective intelligence factor based on their performance, essentially a Team IQ. Indeed, the Team IQ had only the smallest correlation to the average and maximum IQs of the team members. These findings are disconcerting though. If the individual IQ of the team members is not the key determinant of Team IQ, then creating a great team is not really about assembling the smartest people you can find. What matters then?
A model by University of Michigan researchers Lu Hong and Scott Page demonstrates conditions where a more diverse group that is randomly selected can outperform a less diverse group composed of ‘smarter’ people. But this is just a simulation where diversity and IQ are seeded numbers. What about in real life? All the signs still point to diversity. A team of scientists from diverse fields was better able to arrive at solutions to an experimental problem than teams from the same field, even though the solution they sought was within their own experimental framework. Cross-functional teams in companies performed better than teams of only engineers. The reports are many. But is all diversity good?
Diversity can be divisive as much as it can be performance enhancing. In my own experience, teaming people with different education levels can sometimes lead to insecurities and conflict that are more destructive than productive. And adding diversity into a homogeneous group can often be dangerous with the minority ‘diversity’ being spit out of the organization like an organ rejection. So what dimension of diversity really matters? Researchers at the University of Wisconsin looked at the productivity of garment workers in a factory to find clues. A lowly task you might think but indeed the same individuals could collectively produce greater output in teams. And more important, what mattered for team performance was diversity in skills and abilities. On the other hand ethnic and age diversity reduced the team performance, because communication and camaraderie tends to be more difficult across ages and ethnicities. In the MIT collective intelligence study, gender diversity contributed to better team performance. Cognitive diversity enhances Team IQ but many dimensions of demographic diversity detract.
So how do you build a great team? Here are my general rules of thumb, put together people whose skills and experiences are different and therefore think differently. But make sure that they are similar in other respects such as age, education and ethnic background and therefore have things to talk about other than the tasks at hand. If you are selecting by test, the best grouping might be people who have similar scores but who differ largely in the questions or tasks on which they scored their points. And if you can, do it right from the start; seeding diversity in a homogenous grouping is much harder and fraught with perils.
Even as researchers continue to debate how to assess diversity of skills and cognitive styles and their combinations, it is not something that organizations can afford to ignore. As the world changes at increasingly rapid pace, what we need are teams with high IQ for any sort of problem. Recruiting based on the number of years of experience within the same domain and function is increasingly a criterion to do away with.
It’s not just who you hire that matters. Who you put them with may matter far more.