2016 and the Split Prior: An Investigation into the Effect of Social Influence on Crowd Wisdom Predictions of Mass Behavior in Political Contexts
67 Pages Posted: 28 Jan 2019
Date Written: October 13, 2017
The following study is an exploration of the applicability of the wisdom of the crowd phenomenon to the prediction of mass behavior in political contexts. I give a brief review of the crowd wisdom literature and the attendant advancement in usage of the crowd wisdom phenomenon through prediction markets and forecasting tournaments. I further explain and discuss recent advancements in the theory of the crowd wisdom phenomenon and the so-called “surprisingly popular” principle in evaluating crowd wisdom responses. As a case study, I perform an analysis of the outcome of the prediction market called Good Judgment Open (GJO) in reference to the forecasts made in response to the question “Who will win the 2016 US presidential election?” I perform a quantitative analysis of all forecasts made on the 2016 presidential election outcome from the date of October 28th 2016 onward, and a qualitative, discourse analysis of comments made by forecasters in relation to this question over the entire period that the question was “live”. Two related research questions are addressed. Namely, in what way may various social conditions modulate the effectiveness of the wisdom of the crowd phenomenon? And also: Were forecasters who predicted the outcome of the 2016 US presidential election correctly in possession of some special knowledge or forecasting expertise, or were their predictions predicated on something else? I discuss implications for the theory of informational social influence, and present recommendations for future research.
Keywords: Crowd Wisdom, Social Influence, Brier Score,
Suggested Citation: Suggested Citation