Ratan RA, Chung JE, Shen C, Williams D, Poole MS. Schmoozing and Smiting: Trust, Social Institutions, and Communication Patterns in an MMOG. Journal of Computer-Mediated Communication. 2010;16(1):93-114. Available at: http://doi.wiley.com/10.1111/j.1083-6101.2010.01534.x
(I actually got hold of the Word doc format from the Virtual Worlds Observatory, but they don't seem to want me to link directly.)
This is the second paper I've read that uses the same data gathering run. The other one was looking at gender roles, and I used it with the HCT reading group. The only author in common is D. Williams. They managed to get Sony to agree to let them have survey and behavioural data from Everquest, and also got Sony to create a special in-game item as an incentive to get players to fill out a 25 minute survey. Given the size of that opportunity (and how easy it must have been to negotiate) it doesn't surprise me that there is more than one paper as a result - but these are totally different! I would be interested to know how often that happens.
The authors were considering the levels of trust between players, and have a good review of recent changes in social behaviour (e.g. reduced engagement in civic/social events) and how they have changed the opportunities for trust. Prior research has shown that MMOG players list social reasons as important motivators for playing, so they wanted to explore if this social arena leads to new opportunities to trust.
(I need to follow up some of their references on trust and social institutions, particularly around the three major contributing features they outline.)
The paper considers the guild as a smaller social unit than the game itself, which in turn is a smaller unit than the internet at large. (Should also follow up some of their references on guilds!) They then look at trust by asking the players in the survey to rank whether guildmates, others in EQII, and others online "can be trusted" (on a 5-point scale). They did also ask a baseline question about how trusting individuals were ("Generally speaking, would you say that most people in everyday life can be trusted or that you can't be too careful in dealing with people?") They also used age and gender as controls.
They found that (as expected) they did support their hypotheses that players trust their guildmates most, then the other game players, and people online in general least. However, they found that the baseline question was the strongest predictor of the other levels of trust. I think that's interesting - how naturally trusting someone tends to be does show through in their online activities. The writers do seem to have just looked at the mean trust level - I'd like to see if people with lower baseline levels of trust had a smaller range of differences across the three classes, or if they still increased their level of trust by the same amount, just from a lower start point.
The other parts of this paper looked at mode of communication in relation to trust, and whether self-disclosure had any impact. They didn't have access to the content of messages, so had to rely on players accurately filling out the survey on how often they talked about personal matters. Having discussed (briefly) running a guild with a friend, they had rules in their guild that they were not supposed to discuss out-of-game details in raids at least. The paper finds that increased self-disclosure did relate to increased trust, so do guilds with rules similar to that actually risk shooting themselves in the foot as far as trusting each other go?
I was a bit worried when I read this paper that it might negate my line of attack, but I think it could be very useful while leaving me plenty of room to manoeuvre!
(Citation for gender paper: Williams D, Consalvo M, Caplan S, Yee N. Looking for Gender: Gender Roles and Behaviors Among Online Gamers. Journal of Communication. 2009;59(4):700-725. Available at: http://doi.wiley.com/10.1111/j.1460-2466.2009.01453.x [Accessed August 29, 2010].)