I read it as a polite reply to (the hype over) danah boyd's blog post on the differences between MySpace & Facebook: boyd herself never reduced the argument to black vs. whites, but that's the reading everybody had. They had to reply to it, and they wisely avoided to stirr up more flames at the time (their silence was a good pointer that the intuition was onto something, as this study indicates). Now that the result is far less “scandal prone” they closed the question with an interesting approach. But the intended reader was the science-skiming media, hence the rather sound methodolocial base, but the not so academic presentation of the results: with some work, they could have published somethign based on that for, say, ICWSM, but they haven't (so far). Doing so, they also don't have to justify their use of too much, too personal data — the kind no IRB would let any academic handle. By the way, I'd love to have some academic lobbying from you guys to ask Facebook for some run time on that database: they have gold in their hands to answer so many questions — and we are all interested not in detailed data, but the statistical results. If we can negociate that they let us run some scripts, provided those can't possibly reveal any personal information, won't run against their corporate interest, and could help social science, that would be amazing. Maybe some Ivy League universities already pay for that. That offer would lead to lots of interesting discussion on what makes Facebook users different from the general Internet users, way more then what they can have internally—something valuable for the company. About doing the same study abroad: Outside of Nothern America there are no universal, easy to study, significant social class markers; where race carries a similar stigma as it does in the US, such studies are often prohibited. I have no doubt Facebook uses similar approaches for marketing purposes, but any details would be commercially invaluable, and the insights probably as “dull” as what they are in the US after a few years of significant presence.