Does Wisdom of the Crowd Always Prevail?
In 2017, the University of Pennsylvania’s Damon Centola published findings on the affect different network structures can have on the “wisdom of the crowd.” Centola’s experiments display that the idea of “groupthink” happening when people converse is not always entirely accurate.
Participants of the study were given, “estimation challenges, such as guessing the number of calories in a plate of food.” They were then able to observe the guesses of their contacts and change their answer. Finally, they were able to look at the revisions of their contacts and revise once more. Participants were incentivized by cash payment, which was only received if an answer was correct.
Centola found that the network setup leads to different results. Three networks were used: control (no social network), egalitarian, and centralized. In an equal network where everyone has equal number of contacts and equal influence, the “wisdom of the crowd” holds true, as the overall accuracy of the responses increases. However, introducing a centralized networks where participants are connected to one “opinion leader” only can often lead the group astray. Overall, opinion leaders were statistically more likely to lower the accuracy of the group.
In lecture on the 28th, we touched on information-base benefits, which are colloquially referred to as the “wisdom of the crowd”. We saw that once a participant sees others guessing one value, they will follow suit. This is illustrated in the equal network present in Centola’s study. I found it interesting that your available network can greatly affect your decision, as having a very limited network — like the opinion leader in Centola’s setup — is in most scenarios detrimental to your decision-making. Hopefully, readers of this article will be mindful of this finding next time they are faced with a decision.
Article: https://phys.org/news/2017-06-refute-groupthink-wisdom-crowds-prevail.html