Andrew Mao: Studying Teamwork and Cooperation in the Virtual Lab

January 20, 2016


MIT Media Lab, E14-240


For decades, physical behavioral labs have been the primary method for conducting controlled experiments of human behavior. However, the Internet now enables "virtual labs" using online participants, which extend the scope of experimental methods and allow for studies of increasing complexity, size, and breadth compared to physical labs. In this talk, Andrew Mao will discuss two experiments using virtual lab techniques to study interaction and coordination in groups, and motivate the potential of future virtual lab research.
First he will describe a study of collective intelligence in "crisis mapping," where digital volunteers organize to assess and pinpoint damage in the aftermath of humanitarian crises. We use a simulated crisis mapping scenario to evaluate organization and effectiveness in teams of varying size, and find that although individuals in larger teams contribute less effort and work on less demanding tasks, the potential benefits of coordination can outweigh this loss in raw productivity. Mao's team's work shows how such experiments can advance the science of collective intelligence, and motivates further study on the processes underlying digital, decentralized teamwork.
Mao will also present a month-long study of cooperation in prisoner's dilemma using the virtual lab. While cooperation in the real world typically happens over repeated interactions and long periods of time, experiments of cooperation in physical labs are limited to sessions of little more than an hour. His team closes this gap with a long-run experiment of about 100 participants on 20 consecutive weekdays, finding that a group of resilient altruists are able to sustain a high level of cooperation across the entire population. These results shed new and hopeful light on the long-term dynamics of cooperation and also demonstrate the importance of conducting behavioral experiments on longer timescales than previously contemplated.


Andrew Mao is a postdoc at Microsoft Research in NYC in the Computational Social Science group. His work focuses on studying collective intelligence on the Internet, such as teamwork in online communities and coordination in crowdsourcing systems. He specializes in novel experimental techniques on the Internet and in gathering data from real-time, interactive, web-based behavioral experiments. Andrew received his PhD from Harvard University in 2015, where he was advised by Yiling Chen.

Host/Chair: Scalable Cooperation

More Events