The Impact of Social Networks in Wisdom of the Crowds
http://www.pnas.org/content/pnas/early/2017/06/06/1615978114.full.pdf
When determining the validity of information we receive, it is logical that we consider many factors. I frequently rely on the answers to the questions, “Where did the information come from?” and “Is it reliable?” Subsequently, it follows that in order for the source of information to be reputable, it must be created, or at minimum approved, by an authority figure or subject expert. This would usually point to a particular person who has the appropriate qualifications. However, there are other ways in which a document or any source of information can be considered legitimate. This is where the idea of the “wisdom of crowds” comes into play. Since its introduction over 100 years ago, this principle shows that the collective information of a group can provide as equally accurate a result as one expert. This is one process which can be used to attain accurate information without turning to a specialist.
So how do we know that we can truly rely on the wisdom of the crowds? As highlighted in the article above, it is a continual challenge for scientists to understand how network structure affects the wisdom of crowds. In this article, the authors present their finding from an experiment to better understand and model the way in which social influence promotes learning dynamics. Results of the study show that social network structure greatly impacts the wisdom of crowds. In decentralized networks, there was a large drop in the diversity of group estimates. Interestingly, the authors determined that this did not undermine the wisdom of crowds. In decentralized networks the wisdom of crowds actually improves. This is contrary to the generally accepted idea that social influence would undermine the wisdom of crowds. That expectation comes from the idea that social influence will decrease the diversity and an individual’s independence, generating less accurate information. Conversely, the study showed centralized networks experienced group estimates that gravitated toward the belief of a central individual. The article gives an example of this by explaining that if a central node had a higher estimate it led to an increase in the group mean, and similarly if a central node had a low estimate, the group mean was lower.
The most interesting part of this article to me was the author’s ideas for the implications their finding would have on democratic type organizations. They suggest that unlike previous studies, encouraging individuals to communicate during activities like voting could lead to better results. This could greatly change the way companies structure their communication networks by designing processes that encourage collaboration as opposed to trying to keep individual opinions separate. This article provides strong support for the importance of network structure in our society. In fact, the dynamics of our networks are so critical that they can significantly affect our decisions. After reading this article, I feel I have more information, from a reliable source, on how social networks and wisdom of crowds can impact our society.