I think it's easy to misread the data in this case. People fall into two camps, those who draw attention to themselves and those who prefer not to. The latter c

I think it's easy to misread the data in this case. People fall into two camps, those who draw attention to themselves and those who prefer not to. The latter camp is under-represented online precisely because representation would mean drawing attention to themselves. So, counting just the online denizens introduces significant bias in the data.