Eh, I don't think the number of Facebook friends is a meaningful comparison here. At least, not without additional controls. If a user takes the time to seek out such niche likes on Facebook, they also probably take the time to seek out friends to add. I know Facebook doesn't have the massive fake user problem that plagues Twitter, but still, I'd wager that fake profiles, inactive accounts, and users from countries with low Facebook usage skew the average significantly.
You can't really tell if they are "social [media] pariahs" unless you compare them to a representative sample. You could control for the average number of likes, and compare that to the average number of friends. Or even better, you could compare them to several groups of users who like other fringe topics (from all parts of the political spectrum) and see if they're significant outliers. In fact, this is going on my idea list. :)
yea, that seems a better way to do it. and regardless you're dealing with such a small percentage of white nationalists and such a selected sample that it can only be suggestive. but i'm really not sure how else you could get any clues, so it seems worthwhile.
-- seth
You can't really tell if they are "social [media] pariahs" unless you compare them to a representative sample. You could control for the average number of likes, and compare that to the average number of friends. Or even better, you could compare them to several groups of users who like other fringe topics (from all parts of the political spectrum) and see if they're significant outliers. In fact, this is going on my idea list. :)