Ted Castronova will present:

The New Media Petri Dish: Using Virtual Worlds in Social Science Research

The natural sciences have made so much progress in the last 400 years largely due to the method of controlled experimentation. These methods have generally not been available to the social sciences. Social science questions generally live at the level of all society: What happens to crime rates when sentencing guidelines change? How does the Fed's interest rate affect housing supply? What does campaign finance reform do to the political power of marginalized communities? To use a controlled experiment method, one would want to build a series of replica societies and then perform targeted interventions - increasing sentences, interest rates, or campaign contribution limits -- that vary slightly in each replica. Then one would measure the target variable (crime, housing, voting patterns) to see what happened. The problem of course is that we have been unable to create replica societies. Correction: until now, we have been unable to create replica societies. Now, the technology of virtual worlds allows us to make hundreds of versions of pocket societies, each one just like the next. These pocket societies can have dozens or hundreds or thousands of members, and persist for years. Moreover, since they are synthetic from the ground up, they can be altered in precise, controlled ways. Virtual worlds thus may be the technology that allows social science to make progress as rapidly as the natural sciences do - for a similarly great benefit to humankind. This talk discusses the possibilities and pitfalls of this kind of research method, and presents initial results of a proof of concept study.