This project will make use of computer simulation techniques to understand how different ways that scientists make day to day choices might influence the ability of the community of scientists to succeed in learning about the world. Through decades of careful study, we now have a deeper understanding of the various factors that influence choices and beliefs of scientists. While we now understand the causes of scientists' behavior better, we do not have a very detailed understanding how science as a whole is influenced by the small scale choices of individual scientists. Here are some questions of that sort that need to be addressed. Does misconduct on the part of a few scientists have a effect on the reliability of science as a whole? If scientists are motivated by particular political ends, does this color the output of their research?
Intellectual Merit The PI will employ new simulation methods from the physical and social sciences in order to answer three primary questions: What are better and worse social arrangements for science? What is the effect of different ?impure? behaviors on the overall reliability of scientific practice? And what is the effect of efforts to reduce these impure behaviors? Learning the answers to these questions will both help us understand when scientific communities should be trusted and when they should be questioned.
Potential Broader Impacts Beyond its contribution to the scholarly understanding of science, this research will also help to guide science policy so that scientific disciplines might be made better. The PI also plans to develop courses for both graduate and undergraduate students to help further the understanding of scientific practice.
We are told two oddly conflicting things about groups. On the one hand, we are warned about "group think" – that groups behave in stupid ways, doing things that no individual would do on their own. On the other hand we are told about the "wisdom of the crowds" – how groups of uninformed individuals can pool their knowledge and be greater than the sum of their parts. How do we resolve this apparent contradiction? Are groups smart or stupid? What leads them to be good or bad? One very important group, which we might hope to understand better, is the group of scientists. While individuals scientists can be very smart, they can also behave in ways we regard as petty, small minded, and inappropriate. Are these the result of a sort of scientific group think as some have suggested? Or do these small errors wash out in the way that crowds can be wise? This research investigates a number of questions related to this broad research agenda. We have uncovered a number of things that might lead groups of scientists to succumb to a sort of group think – to systematic error – and other things that can make a community of scientists even more brilliant than the individual members. Perhaps one of the most surprising results from our research is how unrelated individual "wisdom" and the wisdom of the crowds can be. In a series of publications we have shown how under the right conditions individually quite stupid scientists can, when placed in a group, outperform individually smarter ones. But, as one might guess, not all groups and not all stupid individuals can benefit in this way. Much of the research on this grant has focused on understanding what features of individuals scientists and features of their community might help to make the best crowds. Our hope is to understand how science can be made more efficient and more productive. In addition to focusing on individual wisdom, we have also looked at how individuals communicate with one another, how they decide whom to trust, and what projects to pursue. We have looked both at how scientific groups should be best structured when engineered from the "top-down," but also how one can create the right sort of incentives so that scientists themselves choose the best structure for their communities from the "bottom up." Making science better will of course improve our lives in unimaginable ways. But, even beyond our research's impact on science policy, our results can have important implications beyond science. Many groups outside of the halls of academia are nonetheless involved in activities like scientists. Friends try to find the best restaurant in town, corporations try to find the right market, and non-profit organizations try to find good volunteers. Understanding how scientists best learn about the world can spill over and help other groups structure themselves in ways that can make them better at what they hope to do.
