Can a person be a crowd?

An intriguing recent area of research has extended the logic of the wisdom of crowds to individuals. It turns out that people can achieve some of the benefit of a crowd by digging deeper into their own minds. The key insight is that people typically rely on only a sample of the evidence available to them at any given time. But what if people had a reset button, so that they could retrieve facts from memory anew or handle the same facts in a new way? Simply asking people to answer again does not work—people will inevitably anchor on their initial opinions. There are at least two effective ways to break this anchoring effect, both illustrated in recent papers in Psychological Science. First, Edward Vul and Hal Pashler showed that people can be freed from their original answer by delaying a second answer. With the time delay people may forget their initial perspectives and think about the problem differently. The second approach, developed by Stefan Herzog and Ralph Hertwig, is to ask people to assume that their first answer was wrong and to answer the question again. Overall, averaging two opinions from the same person using either time delay or ”assume you’re wrong and answer again” improves performance by about half as much as averaging across two people.

That is from Soll, Mannes, and Larrick (forthcoming). The “wisdom of crowds” effect. In H. Pashler (Ed.), Encyclopedia of Mind. Sage Publications. 
The paper is short, worth reading, and also disucsses the psychological obstacles to crowd wisdom.