Wednesday, June 27, 2007

on the hardness of your data

I've been following the comments and discussion on Danah's latest paper about Myspace, facebook and class.. And something is jumping out which is tangential to her (very interesting) points, but which is funny nevertheless.

All these folks are jumping all over the paper because it is apparently unscientific because it either
a) lacks hard data (she's using ethnographic methods), or
b) can be explained by factors she mentions in passing (that have to do with the initial origins of these services)
(there is also the c of people who don't get the difference between describing patterns and predicting individual behavior)

now b actually SUPPORTS HER POINT, which is kinda funny. She is observing phenomena and describing it, her description appears accurate and much of the history of myspace and facebook sees to support it. Somehow some critics think pointing out that history debunks the observation. hee.

and a)-type criticisms are especially funny because many of them sound like criticizing oranges for not being apples.

but also because she's talking about things like class and race and identity and subculture

if anyone can show numerical data on class, it will be rely on a definition of class (or identity, or sucultural boundaries). That definition can NEVER be found by using numbers first, but will have to rely on qualitative categories, which must be derived from some kind of other framework. The more someone relies on numbers in discussing class (especially in America) the more I am suspicious that they aren't serious about understanding what the "class" means.

"50% of all myspace users are working class" would be far less meaningful than the kind of description Danah gives. Because she derives it from actual observation.

It's a perfect case in support of qualitative research actually.

You could have numbers about the number of myspace users who check off "male" or "female" in the gender box, (although even then, what would that mean? what kind of data would you need to make sense of those choices?), but you can't easily have useful numbers about complex but real social institutions. What you have to do first is exactly what this paper does. Observe. Talk to people. look around. broaden your outlook. reach far and wide.

anyway, I was just struck by how funny it was that people were looking to numbers for meaningful statements about class or identity. that's like if science were personality tests. or personality tests were science.

1 comment: