THE Wisdom
of crowds

(Google Scholar Page)

(Research Statement - Download PDF)

One of the most common theories of collective intelligence is that social influence undermines collective judgments through dynamics such as herding and groupthink.  In contrast, my research shows how social information processing can reliably improve belief accuracy—under the right conditions.  

wocbar.png
 

Under Review

Becker, J., Guilbeault, D., & Smith, E.B.  The Crowd Classification Problem. (R&R at Management Science) [read] [data]


Selected Publications (see vita for complete list)

Becker, J., Centola, D., & Porter, E.  (2019). The wisdom of partisan crowds. Proceedings of the National Academy of Sciences, 116(22), 10717-10722. [read] [data]

Centola, D., Becker, J., Brackbill, D., & Baronchelli, A. (2018). Experimental evidence for tipping points in social convention. Science, 360(6393), 1116-1119. [read] [code]

Guilbeault, D. Becker, J. & Centola, D. (2018) Social learning and partisan bias in the interpretation of climate trends. Proceedings of the National Academy of Sciences, 115 (39) [read] [data]

Becker, J., Brackbill, D., & Centola, D. (2017). Network dynamics of social influence in the wisdom of crowds. Proceedings of the National Academy of Sciences, 114 (26), E5070–E5076 [read] [data]


In Preparation

Becker, J. To group or not to group? The network structure of collective intelligence. [draft coming soon]

Becker, J., Almaatouq, A., Horvat, A. The accuracy benefits of group discussion depend on initial belief distribution. [read] [code]

Becker, J.  The network dynamics of social and technological conventions. [read] [code]

Smith, E.B. and Becker, J. Evidence for a collective intelligence theory of market efficiency.

Guilbeault, D., Becker, J., Woolley, S. Nuance matters: Collective intelligence dynamics in the spread of misinformation.

Centola, D., Becker, J., Zhang, J., Guilbeault, D. Social learning improves clinical diagnostic accuracy.

Becker, J. The wisdom of crowds: From collected intelligence to collective intelligence.


Computational methods

Agent Based Modeling                     

By formalizing social theory with computational models, researchers can excel in two key goals.  First, we can use models to determine whether a theoretical account provides sufficient conditions to explain a phenomenon of interest.  Second, we can use models to provide precise theoretical predictions, developing clearly defined hypotheses to be tested with experimental design and observational data.

Tools:  Java, R, NetLogo

 

Web Experiments

Web-based experiments provide unprecedented control over experimental procedures, allowing researchers not only to design every aspect of the subject experience, but also to obtain high-fidelity data that records every interaction by every participant.

Be sure to check out the amazing empirica.ly platform.

Tools:  MTurk API, PhP/Zend, Meteor, ReactJS

 

Network Analysis

Graph theory provides a lingua franca that enables a broad interdisciplinary dialogue among social researchers, physicists, and computer scientists.  While my current research focuses on agent based modeling and experimental design, my coursework and research fellowship has given me in depth experience in observational network analysis including time-series analysis (multiplex networks) for social networks as well as text analysis using semantic networks. 

Tools:  R (igraph), Java (Gephi), Twitter API, Selenium