So you want to study virtual worlds?


Take a look at my papers.

Marina Fiedler and Ernan Haruvy (2009), The Lab versus the Virtual Lab and Virtual Field – An Experimental Investigation of Trust Games with Communication, working paper.
Marina Fiedler, Ernan Haruvy and Sherry Xin Li (2008), Social distance in a virtual world experiment , working paper I hope you will find them informative.

I am currently working on asset market experiments in Second Life and hope to have the paper up soon.
Also, you may find my Powerpoint presentation informative.

Lastly, what I know about experimental economics studies on Second Life is because researchers in the field send me their stuff. So please, please, please, if you are working in this area and have a working paper, an abstract, or a description of what you are doing, please let me know.


Here is more stuff you need to know.

The first economic study of virtual world economies and markets is by Castronova (2001), involving the virtual world of Norrath. At the time, Norrath was the most market-oriented virtual environment. Today, Second Life is the most widely covered virtual world in the business press, with numerous S&P500 companies, including IBM and Sun establishing presence online. Other examples of business applications in virtual worlds are hotels that allow their customers and business partners to walk through the virtual hotel, thus providing useful and inexpensive feedback; clothing companies that enable customers to try out clothes and furniture based on their avatar’s specifications, as well as intercultural sensitivity simulations (Piller and Salvador, 2007).

Over150 educational institutions own land, hold events or collaborate in Second Life. Companies and universities use it to test concepts and designs, conduct work meetings, seminars, lectures, recruiting, advertising and any kind of collaborative activity. It has a functioning and active stock exchange and numerous businesses that sell virtual and real products and services to residents. Second Life users come from all walks of life and are there for various reasons, including socializing and role play. Many of the residents spend a sizable portion of their time in the virtual world, own or rent virtual real estate and many have virtual jobs.


Often when people find out that I study economic behavior in Second Life they have the initial impression that somehow this is not as serious a research topic as some of my other research. They often ask: "How do you play this game?" and "How do you win?"
    Virtual worlds are by definition not games. They are persistent state environments. That means that they exist whether or not you are there, which is different from a game. You can occasionally play games in them and some people do. In some of the more adventure oriented worlds (WoW and Everquest), you can go on adventures. But at the end, you exist in these worlds with all that this entails. You work, socialize, invest, and yes ... play (when you play games, hopefully these are dictator, ultimatum and trust games-- they are fun!).  In Second Life, which is the only world I truly studied, there are relatively few games (although Poker is technically a game). There is a lot of socialization (bars, restaurants, dance clubs, resorts), shopping (lots of shopping malls), sight seeing (you can go to Rome, Paris, Tokyo, ...), investing (there are active stock markets and lots of profitable businesses), running businesses, various interest and discussion groups, many real-world corporations (I recommend visiting the Sun campus) that find this mode of interaction useful, many educational instituions (including, very likely, your own university), and many other useful activities.   It is a mode of interaction and communication that is more fun, more efficient (teleporting is fast), and richer than other modes. So it is not surprising that it is catching on. But reducing it to a game is totally missing the point, and hopefully this little discussion made it clear.

When I make the claim that studying economic games or economic behavior in Second Life gives you access to a  real-world environment over which you have some control, people often get argumentative. Most people agree that the laboratory can be an seen as an artficial setting, thereby distorting behavior, but they don't think that a fantasy world is any more "real" than the lab. You will never convince these people. Don't bother (unless they are refereeing your paper, in which case do bother).  But any of the major virtual world has millions of residents. These residents spend much of their real time and real money there and have true friendships there. To them, this is real, just as Facebook or MySpace are real. If a referee is argumentative you might want to try and make that point somehow.   


Researchers in the field:


Robert Bloomfield is the most passionate and active advocate of virtual world research in experimental economics. He is an active author on Terra Nova which is a blog about virtual worlds. He has written the very influential "Worlds for Study: Invitation - Virtual Worlds for Studying Real-World Business (and Law, and Politics, and Sociology, and....) "  Robert Bloomfield is also the force behind Metanomics , which the most engaging and informative community on Second Life research topics.

Louis Putterman Studied the trust game on Second Life.
Martin Spann has studied the dictator game in WoW

John Duffy has provided the most elegant critique of Second Life experiment, titled Trust in Second Life. You should not run Second Life experiments without reading it.
Thomas Chesney, Swee-Hoon Chuah and Robert Hoffmann wrote a paper titled Virtual World Experimentation: An Exploratory Study which examined dictator, ultimatum, public good, guessing, and minimum effort games in Second Life.

Marina Fiedler is my co-author for most of my Second Life works.

Sascha Fullbrun Studied the trust game on Second Life.

Kevin McCabe His experiments with Peter Twieg and Jaap Weel on Hurricane Island will blow your mind. Very cool.

Interesting and useful books on Virtual Worlds

Byron Reeves is a professor at Stanford University and the author of Total Engagement: Using Games and Virtual Worlds to Change the Way People Work and Businesses Compete And see related blog.


Other common questions/critcisms


1. Do you pay fake money for virtual world experiments? Are the stock markets make-believe?
    No. Linden dollars are convertible to dollars in a real exchange market, with price changing by the second (this is not an exaggeration) and based on demand and supply. As far as I can tell, there is no fake money. Stock markets in Second Life tradeshares in companies that earn real money.

2. Can you run experiments for less money on Second Life?
    Yes you can. Since the average salary on Second Life is lower in dollar terms and since subjects don't travel physical distances to get to your lab, you can get subjects for less money. If you or your students are low on funds you (or they) can get data for cheap. However, if you truly want to compare results to the lab or want confidence in your results, you probably want to pay a comparable amount to the lab in real dollar value. Remember that the field of Experimental Economics is based on real economic incentives. Balance these considerations accordingly.

3. Can subjects lie to you about their identities or misrepresent who they truly are?
    Of course they can. They can also do so in surveys outside of the virtual world, but we still find these surveys useful. We just treat some answers with caution.

4. Can subjects do tricky things they can't or won't do in the lab?
    Yes. Subjcts can try to create multiple avatars and pretend they had never participated before. Subjects can try to contact other subjects through ways not visible to you in a way that will mess up your results. There are a lot of other bad things subjects can do. As you gain experience with running experiments in virtual worlds, you will anticipate many of these things and will devise ways to minimize or eliminate them.  As the field matures, these techniques will become standard. If you want to communicate your own technique on this page, I would be happy to post it here. At the end, this environment involves some loss of control, but I think it is well worth it. 

5.Do people in Second Life engage in inappropriate R-rated behavior?
          People do, but not during the experiment. Second Life restricts such behavior to restricted areas in the World.

If you have additional questions you have encountered please send them to me.