Research on gesture transfers and trends

No, not another survey, but a research report on the spread and adoption of gestures in SL. For lack of a better word, I think it's really about how fads spread (though they don't use that term). Gestures were chosen because they're easy to track, not because there's anything special about them. They looked at over 100K gestures (and over 100K users, though I'm not sure how they calculated that).

The press release from the University of Michigan is at http://www.ns.umich.edu/htdocs/releases/story.php?id=7222. Follow the links to find the full report, which is a ten page PDF file (which I haven't read yet).

They got the data from Linden Labs with identifying names removed.

Yes, but my question is whether they were counting distinct avatars or distinct users.

That's the most fascinating part of this to me (though the contrast between Early Adopters and Influencers is quite interesting, too).

I didn't know LL was willing to do such a thing.

If Linden Lab doesn't have a way of identifying distinct users among avatars, then researchers relying upon that information cannot do so.

The paper does include the methodology, and if you look at the criteria used to select the avatars the researchers tracked, that might answer your underlying questions. Since they used avatars with a demonstrated high degree of social activity, they probably didn't include a lot of bots, dead accounts, and seldom-used alts.