a thoughtful web.
Good ideas and conversation. No ads, no tracking.   Login or Take a Tour!
comment by aidrocsid
aidrocsid  ·  2621 days ago  ·  link  ·    ·  parent  ·  post: Study: Online gaming “losers” are more likely to harass women

Here's the study. I already find the opening of the abstract worrying.

    Gender inequality and sexist behaviour is prevalent in almost all workplaces and rampant in online environments.

That's the beginning of an abstract. A completely unqualified sweeping statement that sexism is ubiquitous. For that matter, the abstract in general sounds as though it's a whole lot of thinking about stuff and not very much actual research. They've done a lot of untested speculation on what their numbers mean right there in the abstract and they admit to setting out to find exactly what they claim to have found.

There are also confounding factors in the very nature of the experiment. Was the researcher playing well? Playing poorly? Were they doing so consistently? We can see that they started out with a very low rank and after only 168 matches managed to get to 50. To me that says that 50 isn't a significantly higher rank than 10. I'm not exactly sure where they're getting their ranking system anyway, as it doesn't seem to resemble Halo 3's multiplayer ranking system. Personally, I'm not a fan of the Halo series, but I do play other FPSes. If 168 matches gets you to a ranking that other players would consider to be "high" then the ranking system is completely meaningless. People who get into FPS games play them frequently for hours at a time, and ranking systems are typically designed to reflect that. So we have someone whose level of actual playing ability we have no idea of and who only ever gets to a rather low skill rank as our "control" as well as both of the factors we'd like to test.

I'm also not sure how not saying anything and thus failing to provoke a response constitutes a control. Especially when they say they only had their coders examine instances in which someone responded to them. So they recorded some control sessions and then just threw them away? This whole thing sounds dicey.

Drawing conclusions from 168 matches, ignoring all confounding factors, no matter how well documented, feels a lot more like using personal experience to argue something you already believed than anything with actual scientific merit.