Some of you might know that I'm a statistician, but I don't believe any of you know that I've been researching the social dynamics of TWW for the past 3 years now.

I initially wanted to look at the organic growth of leadership in a relatively randomized pool of personalities. Unfortunately I got to the game a little late and leaders (a term I use to describe qualitatively as a user who appears to exhibit influence and command respect in relation to other users) had already emerged. Not to mention, the sample of people on TWW is anything but random.

Instead I turned to examine the growth of 'cliques' on the board. It's a basic exercise to classify a users response as either positive, negative or neutral to an earlier users comment. It's also not hard to imagine, graphically, groups of users who are mutually positive towards one another, as well as generally negative towards members of another group. This was done by room (lounge, chit chat, ect) and across all rooms.

As you can imagine, distinct groupings emerged, as well as distinct leaders who generally elicited more positive responses within their groups than any other members. Obviously, there were distinct loners as well (users who invited negative comments from nearly all other users).

Once I established the 'gangs' of TWW, I could look at how these changed over time. For example, you can map how leaders moved in and out of influence within their respective groups.

In particular, for chit chat, I studied the transmission of memes amongst the groups and across the groups.

There are some, in my opinion, very interesting results. I'll post the summary of my article here when I've finished.

2/19/2010 12:51:41 AM

This thread's useless without charts

2/19/2010 12:53:57 AM

sociologists are on the bottom rung of academia, even below education majors

2/19/2010 12:58:00 AM

talk about someone who takes TWW too seriously!

Just kidding, man. I look forward to your charts.

2/19/2010 12:58:54 AM

http://www.chartporn.org

2/19/2010 1:00:01 AM

did you factor in lurkers that randomly post personal stuff and the drama that ensues? also seems like a lot of reading of TWW and possibly missing some things because they are only around for one thread or a few days.

2/19/2010 1:01:02 AM

Oh wow. I really look forward to the followups in this thread. definitely adding it to my topics. I can't wait to see the results 1985!!!

2/19/2010 1:39:31 AM

1985 is probably the most worthwhile poster on tww, but it kind of creeps me out anyone would spend time studying this site

2/19/2010 2:22:49 AM

I'd be curious to read about your methods, and whether they were specific to TWW, or, if you could potentially apply them to other online message boards.

2/19/2010 6:41:15 AM

I hope he tells me I'm popular

2/19/2010 6:48:06 AM

^^

yeah i'd be very interested to hear what kind of methodology you used

2/19/2010 7:02:04 AM

I would like information about said "gangs", their gang "signs and colors", and how to get initiated.

2/19/2010 8:24:55 AM

You’ve been posting here though, doesn’t that bias your observations?

2/19/2010 8:30:10 AM

Quote :

"This thread's useless without charts "

2/19/2010 8:50:31 AM

I'm secretly hoping that my name appears anywhere in this project.

2/19/2010 9:00:13 AM

do i fit in somewhere? i bet not, but it'd be interesting to see. i can't wait to see the results!

2/19/2010 9:41:28 AM

This is just going to serve as an ego-boost to people who probably shouldn't be receiving them

2/19/2010 9:43:35 AM

i imagine the main leaders being EMCE, pilgrimshoes, and Mr. Joshua

for soapbox i suspect theDuke866

[Edited on February 19, 2010 at 10:11 AM. Reason : them duke boys]

2/19/2010 10:10:16 AM

Sorry guys, the actual study was going to be how you responded to this thread, I was hoping to see more posts like IRSeriousCat, predicting what results might look like. But you guys got so excited about this idea, I thought I'd come clean now before I got your hopes up.

Go ahead and lock this thread. Or we can talk about the ethics of performing experiments online without users consent.

2/19/2010 11:10:01 AM

I was just about to post that I thought this thread was the real study.

2/19/2010 11:16:22 AM

hahah funny thread

2/19/2010 11:18:06 AM

Fuck you, 1985.

2/19/2010 11:22:32 AM

god damn y'all are gullible

HEY BROTHER CAN YOU SPARE FIVE BUCKS FOR GAS

I LIVE ON THE OTHER SIDE OF TOWN AND MY TANK IS EMPTY

2/19/2010 11:28:06 AM

I wish you hadn't killed the thread so early. This could have gone into the KG realm of epic threads.

2/19/2010 11:42:56 AM

Yeah, you shouldn't have blown your load so fast

2/19/2010 12:04:33 PM

The more competent thing to do would have been to claim you have all the data you did, and then ASK the user base for their opinions as to who they thought would be the user names at the top of these cliques and gangs.

[Edited on February 19, 2010 at 2:19 PM. Reason : perception vs reality type of question]

2/19/2010 2:18:58 PM

this is disappointing

2/19/2010 2:28:35 PM

You guys assume that the experiment stops when he tells you it is an experiment.

2/19/2010 2:42:51 PM

Free thread! Move this bad boy to chit chat!

2/19/2010 2:59:22 PM

^^ I wonder what the Hawthorne effect will have on this. My null hypothesis is that it has a statistically insignificant effect (p < .01).

2/19/2010 3:55:22 PM

Quote :

"Yeah, you shouldn't have blown your load so fast"

2/19/2010 4:19:57 PM

fail

2/19/2010 4:31:27 PM

Quote :

"wolfpackgrrr: 1985 is probably the most worthwhile poster on tww"

Quote :

"1985: Sorry guys"

Quote :

"Fuck you, 1985."

Quote :

"this is disappointing"

Quote :

"hahah funny thread"

2/19/2010 7:51:02 PM

Quote :

"you shouldn't have blown your load so fast "

2/19/2010 9:38:26 PM

Quote :

"Obviously, there were distinct loners as well (users who invited negative comments from nearly all other users). "

2/20/2010 1:10:44 AM

a comment on your survey is you've forgotten a big part of the population you're studying, and that's the lurkers of a thread. lurkers can also be part of the clique, whereas a member of the clique may be more or less prone to lurk on another members' thread. if you could also track a person's lurking of the thread b4 a response was made (if at all) it may be telling of the personalities...did the lurker emerge finally in response to an opposing clique, or in conjunction w/ a fellow poster?

lurkers that don't log in would be exempt of course...but could also create an error in that portion of the study.

2/20/2010 2:47:46 PM