Social Network Research

You may have read her blog on gaming in The Guardian (UK). Now Aleks Krotoski is doing PhD research on social networks and information diffusion in Second Life, described at http://socialsim.blogspot.com/2006/04/calling-all-second-lifers.html. There’s a short survey to fill out at http://www.psy.surrey.ac.uk/survey21/ and you can visit the Social Simulation Research Lab in the Games sim (205/244/23) (direct SLurl on site).

In a nutshell she asks:

“Who controls the information in virtual worlds? How do fads and trends spread around them? And how important are chance encounters at game parlours, rest areas or marketplaces?

My research aims to answer these questions by describing the social networks of Second Life. The goals are to understand who talks with whom, to follow information as it spreads around the virtual world, and to uncover which groups and cliques are most integral to the social workings of this online space.

I need to poll as many Residents as possible. Who are your friends and acquaintances?”

Aleks has agreed to abide by the privacy policy, and has signed our research agreement, indicating she's working under the Research Ethics policy.

“Who controls the information in virtual worlds?"

I seem to recall that is handled by the Foundation for Information Control.

Just a warning this test takes a long time if you have alot of people on your friendlist.

Wow! I went to the blog and learned something: you can qualify for a PhD focusing on Second Life!?!?! I would love to be on that dissertation committee. Back in the day, I would have loved to have had an advisor broad-minded enough to approve a proposal like that!

This young lady is a pleasure to speak with, unfortunately we were talking when the attack happened on Second Life this evening- just as discrimination became subject of point, Jew decided to make his own...

Ironic isn't it.

how is that the survery goes from social networking to pages on bird flu and such...
that seems rather abstract.

like seriously:
"Does your Second Life Network think it is likely or unlikely that an HIV+ mother will pass on the virus to her foetus in utero?"
wtf is that?

Interesting, but some strange questions perhaps.

The survery doesn't even seem like its about this at all, and I think this survey is being presented under false pretense. This survey seems to be about something else entirely and I'm not sure why this is even being posted with such an obviously false summary.

The survery doesn't even seem like its about this at all, and I think this survey is being presented under false pretense. This survey seems to be about something else entirely and I'm not sure why this is even being posted with such an obviously false summary.

When I got to SARS and Bird Flu I just laughed and quit.

...the ability to network socially in SL? Probably not the most auspicious day to announce that survey.

As for the weird questions (that I didn't get to see because I can't get to my calling card list so had to stop at page two), in defense of the researcher sometimes research design calls for off-the-wall questions for purposes of normalization of answers or a "should I look down this avenue" sort of side-questions.

I don't know the researcher nor have I read the work. I just wanted to point out that there can be legitimate reasons for such "oddities" and given her university affiliation, the research design very likely had to go through a human subjects review panel so one can expect that it is not crazy randomness.

I did note that having a respondent key in their calling card list is patently absurd given that you are asking a person to intermediate a simple data table between two computers that in principle ought to be able to talk directly, but...

No, I really think this survey has nothing to really do with social networking and has some ulterior undisclosed purpose.
1)It asks way too detailed questions about individuals on your list, there is no anonymity and actually gives the reviewer the ability to:
A) collect in world names, this survey is not anonymous
B) Collect names of people who may not wish their names collected
C) be way to specific, surverys aren't that useful when they're about something specific like 47% of people trust Jinsar Eponym, how is that useful information?

2) While the survey starts off interesting about how well you'd trust xxx person and how well you know them it very quickly degenerates into several pages of questions about detailed opinions on SARS, AIDS, and various other health realted questions that:
A) Have no real bearing on anything SL related since I'm pretty sure I can't catch SARS or AIDS through SL.
B) contain way too deatiled scenarios (like what do you think your chance of catching avian flu is when cooking chicken at xxx temperature)

3) The questionaire then moves on to what you THINK your friends think about these types of issues.
A) Again is completely meaningless in the context of SL
B) I can almost guarentee very few people have discussed the above types of issues with their friends on SL and likely would have no opinion on those things.

Hi folks, thanks very much for your feedback. I apologise, I was away and have only just returned to my computer.

I want to assure you that the survey _is_ about social networks in Second Life, but importantly I am expanding on current social network analysis by asking more detailed questions which - it appears - seems strange to many of you. Social Network analysts use surveys like the one available at the in-world research lab and my research website to ascertain network connections between people from their own perspectives. They're called sociometric surveys.

Analysts use various measures (like trust, amount of communication, expertise topics) to create diagrams which depict the social interactions/relationships of people in a large-scale network (in this case, SL). The measures help to determine who's closer to whom on a deeper level than simply that one person knows another. These sociograms result from analysis of the responses of sociometric surveys, and it's through this process that researchers can tell who's important for the control of information through a network.

I have discussed some of the things present in the later parts of the survey with friends, acquaintances and strangers in SL, as well as a variety of other First Life topics. That's why those elements are included in there; they are part of some people's SLs, and I'm keen to understand the depth and breadth of various networks (social as well as knowledge/expert-based ones) which exist online.

When it comes to publication and analysis, I will be using coded numbers assigned to case names in order to ensure anonymity. At no point will names be used in reporting the research, either in official scientific publications or anywhere else where this research may be discussed (e.g., on my research blog). Hope this helps to assure people that their contributions do adhere with international internet research ethics policies (like those from the Association of Internet Researchers: http://www.aoir.org).

I do hope this helps to clarify. I'd like to thank Robin and the others who have helped to publicise this survey. I welcome any comments here or to [email]secondlife@surrey.ac.uk[/email].

Thanks!

Aleks Krotoski (Mynci Gorky)

Your FL topics are too related to come across as those types of questions, as well as too specific.
While you may have come on to SL and initiated discussions about the appropriate cooking temperature of chicken breasts to avoid avian flu, its not somethings thats commonly talked about.

The survey might suit you better if used a variety of FL issues to try and establish trust relationships between individuals, as it is now, I see too much of a pattern for this to provide any kind of useful data. It would be heavily skewed towards those who did discuss those narrow topics and provide no genuine data about people who may be close and trusting yet never chose to discuss each others views on whether or not a mother can pass on AIDs to their child in certain circumstances.

[Tried to send this as a PM but you're not accepting them. Anyway, it still has general value as a post.]


Aleks/Mynci,

After glancing at the opening page of the survey, please understand that I am very unlikely to supply a friends list. Not for the sake of my privacy, but theirs.

Though I may go and give enough data to 'link' me here to my RL name (which happens to be the one I use in UrU), I refuse to give any such info in this Venue to allow that to happen to them. This also happens to fit with my own ethics (in particular that of a long-time UNIX SysAdmin in RL).

I still intend to answer the questions, once I'm back home and rested. Of course, from running surveys myself, I hope you've considered this type of response.

--Alan

Hey, thanks Robin!

Hope it goes well for you Mynci. It looked like a lot of, shall we say, "interesting" survey questions. Wish my browser hadn't crapped out toward the end.

Also, you all should be able to get to the right place by clicking here.


I actually had a whiteshirt drop into the Library earlier and immediately ask where the survey was. If you'd like, Mynci, feel free to put up a big ol' arrow at the entranceway to make it even more obvious.

Hi Folks, I have taken into account the feedback received for the first wave of the survey and am pleased to announce the beginning of Wave 2 of collection using a new and amended survey!

http://www.psy.surrey.ac.uk/survey21

After a massive response from the Second Life community for the first survey released in April 2006, I have released a second survey for the second wave of data collection for research towards my PhD at the University of Surrey. The new survey asks similar questions to the last one - about your social contacts in Second Life - but asks fewer and more specific questions about your relationships in this virtual world.

The new, 14-page Wave 2 survey will take approximately 30 minutes of your time. It involves listing avatars you know socially, choosing drop-down menu options that describe your relationships with them, and selecting answers to questions about your everyday Second Life activities. Please ensure you have access to the names on your Calling Card List whilst completing it.

ANYONE with an account in Second Life can participate! I need to poll as many Residents as possible! Who are your friends and acquaintances? Go to http://www.psy.surrey.ac.uk/survey21/

If you have over 50 avatars on your CC list, I encourage you to contact me directly at [email]secondlife@surrey.ac.uk[/email] for your contribution, and to complete the survey by filling in fewer avatars when appropriate.

All responses will be kept confidential, and Resident anonymity will be preserved at all times. If you would like to find out more about this research, please look around the Research Lab (secondlife://Games1/205/244/23) or check out the research website (http://www.surrey.ac.uk/~psp1ak).

----ABOUT THIS RESEARCH----
My research aims to describe who's connected to whom in Second Life in order to develop a network map of the virtual space. The goals are to understand who talks with whom, to follow information as it spreads around the virtual world, and to uncover which groups and cliques are most integral to the social life of Second Life.

This research has the support of Linden Lab, and is conducted in accordance with the Second Life Research Agreement.

While this has the support of LL and openly declares that the privacy of residents will be protected, I have decided against participating. And I suggest that other residents refrain from participating as well because of the following reasons

i) It does not benefit you or the SL community in any positive way
ii) Your privacy is promised, but it does not stop it from being disseminated via 'gossip'
iii) This is a survey conducted by a former Schutzstaffel member
iv) This will help someone obtain a phd

Mynci:

I received your IM a few days ago and read through your survey. While I find you to be trustworty and your research to be credible and valuable, I too have decided not to participate in your study. In particular, I have a problem inputting a list of avatar names, then telling you who I like and who I do not like, who I trust and who is dishonest, or which people are knowledgeable about things like sex.

I do understand that your research requires you collect names.

But I also hope that you understand that your results will be skewed, even if only a little bit, by the fact that some of us are unwilling to provide this kind of information about our friends.

That being said, I do wish you the best of luck with your survey, and your quest for your Ph.D.

AKA the F.I.C.

Hi all, and thanks very much for taking time to respond. I appreciate your concerns and thank you for bringing them to my attention. In particular, I'd like to address three areas: anonymity/confidentiality, the generation of avatar names and what potential benefit this research may have to the SL community (although I make no claims, just throwing it out there):

1. Anonymity/Confidentiality.
I have been asked by people I have approached for participation in Wave 2 of the SL Social Network survey who referred them to me. I have two answers to this: 1. I can't say because it would breach participants' rights to confidentiality as laid out in the survey and explicity stated in the call for pariticipation and 2. I simply can't without a substantial amount of work, as participants are coded anonymously. Determining such information would require me to sift through an extremely large dataset of avatar names and compare those with an encoded database of respondents.

I have had people contact me very concerned because they have been informed by others that they were the source of the referrals, but I have assured them (to the best of my ability) that this isn't possible for two reasons: 1) the anonymity protocol in place mentioned above and 2) current calls for participation are based upon previous waves of the survey, not the current one. I am not planning on collecting any more data based the information coming in from the current wave of the survey, which began in the second week of June. Furthermore, I was only able to access the Survey Wave 2 responses for the first time this morning.

I apologise for any serious concern this has caused any respondents, and hope that potential participants understand that I cannot tell them who they were referred by.

2. Generation of Avatar Names
As some have indicated, the development of a social network for SL is reliant upon generating a list of names whom respondents are friends with, and that's why it's such a large part of this research survey. If for whatever reason you don't feel comfortable doing this, please feel free to terminate your participation! The survey may not be for everyone!

So everything is clear, the outcome of this research is to have a map of communication patterns throughout the SL virtual community, and to do this I am relying upon existing social network methods for what is called ego-centric network social network analysis. This is usually conducted using sociometric surveys, i.e., by asking participants to list names of their friends and acquaintances, and rating their patterns of communication in order to determine how relatively "close" or "distant" they are from one another. In the end, this information is fed through an analysis programme and a social network diagram is output which visually describes the network of relationships in SL.

If you are curious about other social network research and methods, please explore the International Network for Social Network Analysis homepage, an association of which I am a member (http://www.insna.org/).

3. Possible benefits to the SL community
Social networks are interesting tools used to understand the pathways of communication - and potential gaps in knowledge - throughout a community. My purpose in doing such an analysis in SL is to understand how this process occurs in a unique virtual community, and to potentially inform people of how they may best get messages to different types of communities in-world, to help bridge knowledge-gaps, to help forge connections between other like-minded people, etc. As SL expands in population exponeitally, it can be difficult to get a hold on the enormity of the community. Sometimes patterns emerge from social network diagrams which inform observers where and when social interaction is best achieved, where community is best-formed, and those groups which are best-able to engender community.

I make no claims that my current research will address all of these issues, but I hope that by mapping the social interaction patterns of SL I can contribute something to the community which benefits it in the long-term.

For more information on what social network research has contributed to other communities, here are a few links to SNA journals:
Connections: http://www.insna.org/indexConnect.html
Social Networks: http://www.insna.org/INSNA/sn_inf.html
Journal of Social Structure: http://www.cmu.edu/joss/

If you have any questions or concerns, please don't hesitate to get in touch via IM in-world, or via email : [email]secondlife@surrey.ac.uk[/email]

Thanks very much!

Aleks Krotoski/Mynci Gorky
http://www.surrey.ac.uk/~psp1ak

Now that you've cleared that all up it sounds like data collection to sell to marketing companies who want to target SL.

yes, the test takes very long time, ver Y

Hi Jinsar, I promise I have no intentions or desires to do that.

lou, I'm not sure when you completed the survey, but the current one has radically changed from earlier versions and should take a fraction of the time that the last one did. I base this contention on the feedback I have received from people who have taken it within the past week and a half.

Aleks/Mynci

Please.

Without an email address as part of the survey, why would they bother? After all, there are plenty of avatar databases they could use instead... (and being on those lists, I have yet to receive unsolicited spam)

Further, marketing demographics could be better bought from the mother company (in theory), or through a sales vendor.


To the best of my knowledge, Mynci has been very professional with this survey, addressing my own concerns in quick order. And this is opt-in, as these things should be. If you do not wish to participate, you should not do so.

Personally, I support opt-in studies in Second Life. This one is a bit difficult because it requires some degree of confidentiality and good faith (because contacts are involved), but thusfar the study divulges very little that couldn't simply be accessed using avatar find or a Name2Key database.


-----

Looking at it again, however, some of the blanks (like demographics) should have an "I choose to not answer this question" option.