I’ve written about the benefits of diversity in group performance in the past, as in Diversity leads to better decisions but not for the reasons you’d think, but those benefits were based on observing groups actually interacting: a psychosocial analysis.

In that earlier piece, I reported on findings of research from Katherine Phillips, Katie Liljenquist, and Margaret Neale, who determined that the social awkwardness that arises from socially dissimilar group members are brought into a group setting leads to greater effort in understanding and evaluating alternative perspectives. That effort leads many group members — when the diversity is extreme — to believe that it’s not worth the pain involved, despite the group effectiveness.

There is also a great deal of evidence to support the value of active dissent (see Dissensus, not consensus, is the shorter but steeper path), which is in many ways a way to achieve the effect of diversity artificially. It counters the natural drift toward groupthink in organizations to have individuals avoid the well-worn paths of agreement and the cognitive biases that lead us to pay more attention to those ideas we share and to devalue those that few or none are espousing.

Perhaps the most difficult step in embracing dissent is to drop the tendency to ‘get everyone on the same page’, or use early opportunities to consider the widest possible range of alternatives to instead eliminate ideas. Likewise, companies that create a monoculture at work — like the investment banking firm that hires only Ivy League jocks, or an engineering company where 85% of the senior team all worked 20+ years at GE —  are intentionally minimizing the pain of diversity at the expense of higher results. So the next time you hear about a startup that is hiring people that ‘fit’ the culture, be aware that this conceals (slightly) an anti-diverse HR campaign.

So, companies would be wise to seek wide diversity or to emulate it through active and intentional dissent.

New evidence about the theoretical benefits of diversity in groups has arisen. Lu Hong and Scott Page published a paper, Groups of diverse problem solvers can outperform groups of high-ability problem solvers, in which they present a mathematical model that seems to support the conclusion of the article’s title. The authors are less concerned with demographic diversity (which they refer to as identity diversity), and are focused on functional diversity, meaning the way they approach problems and solve them:

Building on earlier ideas from the psychology and artificial intelligence literatures (7), we describe a mathematical framework for modeling problem solvers that captures the functional diversity that cognitive psychologists and organizational theorists claim is correlated with identity diversity. In our framework, agents possess internal representations of problems, which we call perspectives, and algorithms that they use to locate solutions, which we call heuristics. Together, a perspective-heuristic pair creates a mapping from the space of possible solutions to itself. A diverse group is one whose agents’ mappings are diverse.

They suggest a scenario, in which a company needs to solve a hard problem. They test 1,000 candidates to determine their individual aptitude for solving such a problem. The scores are distributed from 60% to 90%. What should the company do? Should it 1/ hire the person with the highest score, 2/ 20 people with the 20 highest scores, or 3/ 20 random candidates? Past research suggests that alternative 2 is better than 1, because adding more people to a group means that they have a greater range of perspective-heuristic pairs. But past research — the authors argue —  says little about 2 versus 3, although other work — like Phillips and company’s findings — suggests that a broader range of diversity would be best. And the likelihood is that those scoring highest in the test will tend toward less diversity.

The remainder of the article is a dense mathematical proof that shows that, in general, ‘a random group of intelligent problem solvers will outperform a group of the best problem solvers’. And they continue with this great insight:

A further implication of our result is that, in a problem-solving context, a person’s value depends on her ability to improve the collective decision. A person’s expected contribution is contextual, depending on the perspectives and heuristics of others who work on the problem. The diversity of an agent’s problem-solving approach, as embedded in her perspective-heuristic pair, relative to the other problem solvers is an important predictor of her value and may be more relevant than her ability to solve the problem on her own. Thus, even if we were to accept the claim that IQ tests, Scholastic Aptitude Test scores, and college grades predict individual problem-solving ability, they may not be as important in determining a person’s potential contribution as a problem solver as would be measures of how differently that person thinks.

So, the decision of who to add to an existing group to make its problem solving capabilities higher is clear: pick someone unlike the others, but who has a reasonably good skill set — but not necessarily the highest — for the class of problem the group is tasked with solving. The point of diversity is to increase the range of alternative ways to wrestle with the problem space, and so a person with a unique background or set of perspectives is much more valuable than yet-another-math-whiz from Princeton.