Here is my case study of MeFi for an ecommunities class.

April 23, 2005 7:22 AM

Some of you may remember that I've been working on a case study about MeFi for an ecommunities class. Well, as promised, here is my paper.

(As a reminder, this is NOT research.)
posted by nalihasan to MetaFilter-related at 7:22 AM (49 comments total)

thanks. really interesting. i now see that what i thought was wanking is actually reification! :o)
posted by andrew cooke at 7:40 AM on April 23, 2005

(I went straight to the end of the paper looking for shoutouts.)
posted by planetkyoto at 7:46 AM on April 23, 2005

looks great! i like the pretty colours.
posted by hugsnkisses at 7:48 AM on April 23, 2005

Metafilter: a form of "Walt Disney creativity"
Heh.

[well done nalihasan - It's well written and easy to read]
posted by peacay at 8:06 AM on April 23, 2005

Thanks! I hope I represented MeFi and MeFites well. It was a lot of fun to spend so much time on MeFi - for a class project! :)
posted by nalihasan at 8:18 AM on April 23, 2005

Corrections:

Typo on pg 4 under Methods: "Over the course of this past semester, I studied MetaFilter by observing pubic posts".

Also, on pg 10, in the footnotes, you claim, "Deleting a thread or comment removes it completely from the website’s database and archives.". Not true for threads in the blue, which are accessible by entering the URL of that thread.
posted by Gyan at 8:44 AM on April 23, 2005

Typo on pg 4 under Methods: "Over the course of this past semester, I studied MetaFilter by observing pubic posts".

What's a typo there? God knows I've also observed many pubic posts...although with the occasional goatse outbursts, I've seen many more rectal than pubic posts.
posted by ChrisTN at 8:50 AM on April 23, 2005

Don't mind Gyan he is an engineer.
posted by MLIS at 8:50 AM on April 23, 2005

There's going to be another section with highlights about the reactions you get here, right? Please?
posted by ferociouskitty at 9:05 AM on April 23, 2005

*scratches pubes*
posted by quonsar at 9:07 AM on April 23, 2005

Nevertheless, the number of comments that a post receives is usually a good indicator of how the community feels about the post, where posts with a high number of comments are generally considered good. In MeFi's case, a good post is one that can keep the conversation going by prompting members to keep posting comments and replying to one another.

I dont think this is correct. A huge number of Metatalk discussion has been around how it is all too easy to generate lots of comments (post something controversial and/or politically charged) but that this is not the same as the "quality" of a post. Both good and bad posts gets lots of comments. Both great and terrible posts get little or no comments.

Anyways, I'm not drawing a conclusion so much as pointing out that the statement " where posts with a high number of comments are generally considered good." does not reflect to me a valid conclusion.
posted by vacapinta at 9:09 AM on April 23, 2005

I wish to agree with vacapinta. In my opinion, and as a gross generalization, the best posts are the ones that get a handful of 'this is good' and the worst posts are the ones that get 200 comments worth of backbiting and sniping.
posted by jacquilynne at 9:18 AM on April 23, 2005

I kind of agree with vacapinta. Quite often a post about politics with many comments will be very argumentative and abrasive, and reflective of differences expressed ; whereas a non-contentious post - about the life of an artist, for instance - can often have less than 10 comments.

It would be interesting to select a random week/month from Mefi, rank the threads by length, and code them according to type and subject matter (e.g. news, politics, photography, music, etc.).
posted by carter at 9:22 AM on April 23, 2005

Depends what you mean by "good," vacapinta. Its just like other forms of communication. Hollywood blockbusters like "Titanic" get millions of people to plunk down ten bucks to see them and win academy awards; little indie flicks don't. Single link news-filtery FPPs generate a lot of responses; thoroughly-researched ones on obscure topics that don't appeal to the lowest common denominator don't. Which is "good?"

I tend to favor the latter in both cases, but I know I'm in the distinct minority - on the first example at least.
posted by googly at 9:23 AM on April 23, 2005

I think it's just different kinds of good. I rather enjoy the clusterfucks--especially Days of Our MeTa. I also enjoy interesting noncontentious links sometimes. Generally it's people who can't see the entertainment or learning value to conflict who don't really see the goodness of the former.
posted by dame at 9:27 AM on April 23, 2005

I agree with vacapinta: not necessarily that the number of posts does or does not reflect whether the post is good or not, but that the number of posts is not necessarily a good indicator of what the community feels is the quality of the post.

In this discussion alone, we're disagreeing about whether or not the number of responses indicates quality. It can't be that great a barometer of community opinion if the community doesn't generally agree on it.
posted by bugbread at 9:34 AM on April 23, 2005

Depends what you mean by "good," vacapinta. Its just like other forms of communication

Thats my point. I dont intend to start a discussion here. i was merely pointing out that this is a controversial topic but that in the paper it reads as a sort of definite conclusion that more comments=good.
posted by vacapinta at 9:35 AM on April 23, 2005

Single link news-filtery FPPs generate a lot of responses

Why?

For the same reason that when we talk about other people's problems, we're all experts. But when it comes to dealing with internal issues, it's so complicated that a quick fix is impossible. The reason why AskMeFi is so popular, along with NewsFilter, Favorite Band threads, etc., is because it's so easy to have a basic opinion about anything and think of yourself as well versed. And there's no shortage of know-it-alls in the world, myself included of course.

The longer I hang out here, the more I've learned to appreciate the quiet posts that don't generate much overt attention. I learn a lot more when I don't have dozens of people yelling to be heard, making snarky one-liners, and effectively trying to show-off their intellectual prowess. Subtlety is over-rated, especially on the internet where it's easy to be overlooked.
posted by SeizeTheDay at 9:45 AM on April 23, 2005

good paper--thanks for letting us see the final result--I really like the Public Goods, Pro-Social Behavior, and Free Riding & the Community of Practice sections (and the Mutual Engagement part)

Was this a mistake?
There are many grumblings from elders and regulars about newbies posting invaluable links to the front page.
invaluable? don't people grumble when they see a link as the opposite?

And investigating earlier things like the Kaycee Nicole, We Have Cameras, Vibrating Brooms, etc, might have helped flesh out examples. A lot of yours are very recent.
posted by amberglow at 9:58 AM on April 23, 2005

Nevertheless, the number of comments that a post receives is usually a good indicator of how the community feels about the post, where posts with a high number of comments are generally considered good. In MeFi's case, a good post is one that can keep the conversation going by prompting members to keep posting comments and replying to one another.

then, Postroad's front page posts and ParisParamus' comments are the best things on MetaFilter.
interesting thesis.
posted by matteo at 10:14 AM on April 23, 2005

and 9622 is of course the best thread ever
posted by matteo at 10:14 AM on April 23, 2005

To me, more responses usually = least interesting post. My favorite FPP seem to get between 2 and 20 posts. The ones with dozens or hundreds of replies might be interesting discussion, but they're usually links to news or other things that *can* be discussed. There's not much point discussing a link to a great flash page or art page. You either get it or you don't.
posted by dobbs at 10:19 AM on April 23, 2005

Thanks for the feedback, everybody!

OMG - that's a really bad typo! I just changed it, thanks for catching it!

I really don't want to get into a discussion about the relationship between the number of comments and the quality of FPPs (because it really isn't very clearcut and there are so many exceptions) but I think I sort of based that statement on the link guidelines.

We talked a lot about Slashdot in our class and I always find it interesting how MeFi works even without a "reputation system." I guess I was probably wondering if the number of comments sort of acts like a reputation system. I know I always wonder about threads that have a lot of comments - what's all the discussion about?
posted by nalihasan at 10:20 AM on April 23, 2005

well, either the (usually hot-button) issue being discussed has lots of people who care about it one way or the other, and/or someone says something that ignites a flood of responses, and/or more info and background links are added, and/or it's a developing story so keeps getting comments...

the "invaluable"?
posted by amberglow at 10:47 AM on April 23, 2005

nalihasan, it's a pretty good overview, especially for someone that only viewed the site for a couple months. Many early examples could prop up your points, but they were years ago.

On the comments point, I had the same initial thought as vacapinta when I read the paper, but if one has to make generalizations in a paper, I understand how the number of comments has a relationship with popularity and even looser relationship with quality. It's not correct nor incorrect in all cases, but what makes for a good post is so nebulous that it's tough to define in a short overview paper, so I understand why you made the point.

Regarding Slashdot vs. MetaFilter, I've talked a lot about the differences with people lately and it all comes down to technological means of providing checks and balances and social means of providing checks and balances.

MetaFilter runs on a tank filled with guilt and social pressure. If you fuck up, people will react and persuade you strongly to stop. Slashdot runs on a bunch of code, so if you do something someone doesn't like, they vote against you in the interface.

There are a lot of large successful communities where social pressure is enough in the culture to keep things running smoothly, though it's tough to do and requires constant tending.
posted by mathowie at 10:52 AM on April 23, 2005

Although I agree with vacapinta in that the number of comments on a post on MeFi does not necessarily indicate a post or thread of quality, overall I'd have to say that this is a very interesting case study. Well done, nalihasan!
posted by Lynsey at 11:32 AM on April 23, 2005

I just have to give nalihasan credit for putting her paper out there for us to critique.
posted by ..ooOOoo....ooOOoo.. at 12:01 PM on April 23, 2005

OMG - that's a really bad typo!

That's amusing but doesn't impede understanding, since the correct reading is obvious. The misuse of "invaluable" in this sentence is much worse:
There are many grumblings from elders and regulars about newbies posting invaluable links to the front page.
You should change it to "worthless" or whatever you meant to say.

I have to say that (MetaFilter 2005) strikes me as a pretty useless reference; shouldn't you reference statements to individual threads?
posted by languagehat at 1:11 PM on April 23, 2005

Very nice. One for the time capsule!
posted by scarabic at 1:24 PM on April 23, 2005

It's strange to see things like "farkiness" explained in footnotes. Strange and amusing. I can imagine some cantankerous old professors peering at each other through their bifocals and working out the language roots of all this new internet slang.

Cheers, nalihasan!
posted by cmyk at 1:32 PM on April 23, 2005

thanks for noticing my "invaluable" mistake, it's been corrected!
posted by nalihasan at 2:00 PM on April 23, 2005

Well done, Nalihasan. I have nothing pithy to add, no typos to correct (I spotted a few, but was too lazy to note pages -- but look for a stray "o" and a "boost" where you meant "boast"), and no argument with your conclusions. Just wanted to let you know I enjoyed your paper.
posted by mmahaffie at 2:19 PM on April 23, 2005

thanks for noticing the other typos!
posted by nalihasan at 2:51 PM on April 23, 2005

Nalihasan - In your Conclusion, did you mean to say that MetaFlter is currently "boasting" rather than "boosting" more than 20,00 members? On preview, it has been spotted
posted by Cranberry at 2:56 PM on April 23, 2005

Nicely done, and thanks for posting the paper! I really appreciated your comments and observations about Matt's persona/presence/god-like moderator role in the community.

I had another question that wasn't part of your project, but I was curious if you had any thoughts/observations about gender roles/issues. I've gotten the sense that gender issues have a lot of charge here, and I remember Matt once wrote about trying to rein in the "sausage fest" when the testosterone got too wild (sexist posts, comments, aggressive call-outs, whatever). Did you have any thoughts about how gender issues play themselves out here and if it's any different than any other blog?
posted by jasper411 at 3:02 PM on April 23, 2005

Metafilter: Running on a tank filled with guilt and social pressure
posted by graventy at 3:10 PM on April 23, 2005

Interesting paper, thanks so much for posting!
posted by ThePinkSuperhero at 4:58 PM on April 23, 2005 [1 favorite]

Hmm, gender issues would have been interesting to study but I didn't even think about it. Now that I think about it, we didn't even read any papers about gender issues in online communities.

We did read a paper about race issues online but I didn't see that pop up on MeFi.

I think gender issues could be a semester long project on its own.
posted by nalihasan at 5:52 PM on April 23, 2005

I agree with the number of comments not necessarily being an indicator of a great post but it does show some success. Nice paper and thanks for putting it out for us to see.

I also liked that page 13 shows your comment in the poopy thread. It was one of our better more human moments.

and 9622 is of course the best thread ever
MeTa 8762 is pretty good.
posted by geekyguy at 6:04 PM on April 23, 2005

nalihasan, you should submit this paper to First Monday. It's right up their alley.

Also, I think you should send a copy to Robert Axelrod, he'd like it.
posted by warbaby at 6:22 PM on April 23, 2005

We talked a lot about Slashdot in our class and I always find it interesting how MeFi works even without a "reputation system."

I was going to say that there is definitely a reputation system. It's just not a button (or whatever) like Slashdot apparently uses. You only learn about member reputations and/or form one yourself by observing the site.

And then mathowie said something that was better than I was going to say said:

Regarding Slashdot vs. MetaFilter, I've talked a lot about the differences with people lately and it all comes down to technological means of providing checks and balances and social means of providing checks and balances.

I can't really recall the race of MeFi members ever being an issue. Unless you go to meetups, or see photographs from said meetups, you I don't even think about it. There are several aspects of MeFi members that would have been interesting to address (the "boyzone" issues, sexual orientation and religion, just to name a few), but you're the one that knows what's appropriate for your project.

All-in-all a well done job, nalihasan! It's all stuff I "know" about MetaFilter, at least subconciously, but to have it written out point by point was quite revealing.
posted by deborah at 8:14 PM on April 23, 2005

I liked your work. The font and color choices were excellent and enhanced the papers readability.

One thing though: Posts with 250 comments. Who has time to read each comment and really invest any thought to what each comment is about? I think having a bunch of comments creates an apathetic condition for members late to the thread. I vote to call this Meta-Apathy.
posted by Dean Keaton at 2:48 AM on April 24, 2005

We talked a lot about Slashdot in our class and I always find it interesting how MeFi works even without a "reputation system."

While membership was closed, Metafilter really did have a "reputation system" that manifested itself in ugly ways : better-known ("Mefi famous?") members getting away with trolling and really saying quite outrageous stuff. Also the posting history legacy would often precede each member so that even if somebody came up with a valid point, someone else would say something dumb like "Oh, you're that fucked up (insert axe-grinding reason) aren't you? All your opinions are useless".

I've seen less of that now that there are too many new voices to shoutdown.
posted by FieldingGoodney at 2:56 AM on April 24, 2005

Oh, fuck off.
posted by stavrosthewonderchicken at 7:46 AM on April 24, 2005

warbaby - do you happen to know Professor Axelrod? I guess I feel kind of weird just emailing him an unsolicited paper! :)

Thanks for the First Monday suggestion, too!
posted by nalihasan at 9:51 AM on April 24, 2005

I am a bit annoyed that the paper speaks of the open memberships policy as if it has always been that way. The fact that memberships have been closed for the majority of metafilter's history has had a tremendous influence on its development. And that's true whether you're glad open memberships are now on indefinitely, or not.
posted by bingo at 1:07 PM on April 24, 2005

i thought it mentioned that? (although i read it a couple of days and several time zones away).
posted by andrew cooke at 1:56 PM on April 24, 2005

Friend of a friend (also on the faculty). Nice guy. I think he'll like your article.

Have your prof introduce you, if that seems a better way.
posted by warbaby at 2:11 PM on April 24, 2005

Heh, when I saw the (Blood 2000) reference (v. early in the paper before I 'got' your reference system) for when 'blog' became popular I got intruigued.

When I found out that it was just s/o's last name and not the journal... mein gotte, I need some time off.
posted by PurplePorpoise at 8:02 PM on April 24, 2005