Powered by LiveJournal.com
Okay, I hate myspace, but I hate being bad at anything worse than I hate myspace. Therefore, any of you with a myspace account
are encouraged to "friend" me. That way, I can have more than my measly, pathetic, 14 friends. Yes, I really do want to be one of those people with, like, a billion friends. That way, I'll be cool. I'll say "yes" to any friend request.
Pleeeeeeeeeeeease be my friend.
P.s: I am painfully aware that this will not actually affect my coolness in any way. However, it's what I've got. Work with me here.
, what sort of effect will that have on my neat little social network? Huh? Huh?
Done, although I hate myspace as well. You know, I had assumed from your last name that you were Asian or something...
|Date:||May 24th, 2006 01:54 am (UTC)|| |
You know, I had assumed from your last name that you were Asian or something
It's happened before. Think "Duane" passing through Ellis island.
There's another possibility: "Dwan" translates from the old Irish as "short, dark one." This maps well to the time when the name first appears: around 1500. In 1492, in addition to sponsoring a land grab, queen Isabella and her husband Ferdinand kicked the Moors out of Spain. All those short, dark people went elsewhere, many of them to Ireland.
My little sister has a story that takes the cake though: One of her instructors was calling attendance. "Dawn," "Dawn," "Carol Dawn." Finally she said "I'm Carol DWAN." This elicited a glare and a response, "Dawn, please speak up in a timely fashion next time. You should recognize your own name."
|Date:||May 24th, 2006 02:01 am (UTC)|| |
Heh. My wife's maiden name is Newlon, and when she was growing up her teachers were forever crossing the L for her, like she couldn't spell her own name.
Funny thing is, from my last name, some people think I'm Irish. (I'm Taiwanese)
*tries to parse last name*
|Date:||May 24th, 2006 02:12 am (UTC)|| |
what sort of effect will that have on my neat little social network?
I realize that you didn't really intend this question to be taken seriously, but your question isn't really meaningful as stated. If you'd like me to try to answer it (basically by suggesting various things that you might have meant by that and answering them instead), I will, but it might go better over the phone. :)
|Date:||May 24th, 2006 02:17 am (UTC)|| |
What I really meant was:
Does the fact that I deliberately mess with my social network in order to emphasize certain characteristics invalidate the significance of your analysis of those same characteristics.
Or, more broadly:
Once a system is aware that it's being analyzed, how can you possibly analyze it?
|Date:||May 24th, 2006 06:54 am (UTC)|| |
Oh, OK. That I can at least try to answer.
First: what is considered "your social network" is itself open to interpretation. By implication in this case you assert that the link A->B exists if A lists B as a friend, but even in the context of myspace alone you could define other networks.
(So depending on my choice of data, your plugging for friends may have no effect.)
Second: fundamentally, when you're doing social network analysis and are preparing to collect your data, the first thing you do is decide what questions you're trying to ask--what kinds of analysis are of interest. This will suggest something about what data you want.
(If I'm trying to infer who the terrorists are via social network analysis, myspace networks are probably mostly irrelevant.)
As a side point, there are different schools of thought regarding the interpretation of social network data, but it's generally a safe assumption that in any social network analysis, there is missing or incorrect data: not all data was collected, or some data is deliberately concealed, or the data is misinterpreted, or identity confusion (polysemy or synonymy), or you didn't collect data on all the relevant types of connections (almost inevitable).
I prefer to work with network data that's more objectively measureable, e.g., A->B means that A emailed B at least 3 times. You can still argue about what it _means_ that A emailed B > 3 times (if that's relevant to your analysis), but it's much less fuzzy than "A considers B to be a friend".
The problem that you bring up is of course not unique to social network analysis; in _any_ context in which you're looking at human behavior, you have to ask whether you think it's likely that the people being studied are aware of the observation, or are for any reason trying to game the system.
Presumably you're not the only one who's trying to collect lots of friends on myspace as a prestige thing, or as a game. Any analysis of the myspace friend network should take account of this fact. One thing you can do with a model I created a few years ago is to model the strength of a connection according to such things as the similarity of link patterns: e.g., if you and a 'friend' share a lot of 'friends' (esp. if each of those friends has few friends of their own) then it's more likely that the 'friend' relation between you two reflects a real relationship. (This is one example of a feature; there are many others.)
How's that? :)
Friends! Collect 'em all!
If I ever cave for my brother's sake and sign up, I'm in.