If you had a challenging problem to solve, would you trust the wisdom of a group or the wisdom of a crowd? Before you answer, consider this. Asking a large number of people a question and taking the statistical average has been shown to be incredibly accurate. It has also been shown that group decisions can be catastrophic. For example, we previously discussed how to avoid bad decisions by groupthink. Now, Cass Sunstein and Reid Hastie have expanded on this research and written a book on making better decisions entitled “Wiser: Getting Beyond Groupthink to Make Groups Smarter.”
Sunstein and Hastie have identified four common problems experienced by deliberating groups. First, they contend that groups amplify the errors of their individual members. Second, they argue that groups are susceptible to cascade effects as individuals are influenced by those members who speak or act before them. Third, they found that groups can become more polarized by adopting positions more consistent with the pre-deliberation tendencies of their members. And, fourth, they believe that groups focus on shared information, or what they describe as what everybody knows, at the expense of unshared information.
In regard to amplifying errors, the authors maintain that due to people’s predispositions and biases groups can actually make worse decisions than individuals. For instance, the planning fallacy and the biases of availability, representativeness and framing can be amplified in group settings and strengthen people’s convictions.
Cascade effects are largely the result of social influence and have been demonstrated in numerous studies. In fact, we shed light on social influence earlier by citing several ways in which this process manifests itself. Moreover, we commonly see cascade effects when we decide which music to download or which posts or videos to like as we are largely influenced by the choices that others made before us.
According to Sunstein and Hastie, group polarization has been found in hundreds studies in more than a dozen countries. To explore this concept, the authors conducted an experiment among two groups in Colorado. The first was a left-leaning group from Boulder and the second was a right-leaning group from Colorado Springs. They asked each group member to record their views individually and anonymously both pre- and post-deliberation.
The left-leaning groups considered traditionally liberal issues and the right-leaning groups considered traditionally conservation issues. What they found was that each group ended up adopting more extreme positions both as a group and individually than they held earlier. In addition, the authors found that group deliberation increased consensus and decreased diverse viewpoints.
The fourth condition consists of focusing on shared information rather than unshared information. This unshared information can create what is known as hidden profiles, which is accurate information that the group possesses but fails to surface. These so-called hidden profiles are the result of the common-knowledge effect. This occurs when information that is shared by all group members has more of an influence on group judgments than information that is only shared by a few.
In response to these problems, Sunstein and Hastie recommend eight ways to improve group decision-making. The first way is to have inquisitive and self-silencing leaders. The second way is for leaders to “prime” or encourage critical thinking. Third, leaders can reward group success, thereby further incenting individuals to disclose what they know. Fourth, leaders could establish roles for individuals so others are able to identify and recognize them for their expertise.
A fifth way involves perspective-changing for leaders. This essentially requires leaders to ask themselves what new leaders would do. Sixth, leaders could ensure that their groups include authentic devil’s advocates or people who are known to hold different perspectives. Seventh, leaders could adopt so-called red teams, which are typically used in the military and whose function is to thwart a primary team’s mission.
The eighth way entails using a variation of the Delphi method. This method consists of having group members first state their views anonymously and independently, and then sharing those views with others. This process is then repeated until the group converges on their votes. This differs from social averaging as group members can be influenced by the prior votes of their colleagues.
Lastly, the authors recommend a two-stage process for group deliberation – identification and selection. The identification stage is dedicated to generating as many solutions as possible while the selection stage is devoted to critically evaluating the best ideas. This process ensures that diverse solutions emerge and that each one is carefully considered.
Understanding human behavior and group dynamics can dramatically improve any organization’s decision-making. This is especially true for organizations like ours that are in the business of strategy. Therefore, if you want to ensure that strategy is everyone’s business, you need to create the right conditions so that the best decisions can be made.
Owen Eagan is a Vice President for The Saint Consulting Group, an international management consulting firm specializing in land use politics. He is also an adjunct faculty member at Emerson College, the nation’s only four-year institution dedicated exclusively to communications and the performing arts. Email Eagan@tscg.biz