Danah Boyd is a researcher in social media at Microsoft Research. She describes herself as an ethnographer, which literally translates as someone who writes about people. Like other members of her trade, Ms Boyd uses statistical methods as a fundamental part of her investigations.

(Authors also write about people, with the likes of Jane Austen, Charles Dickens and Anton Chekhov being expert people analysts, but they write about people in the particular, not the mass.)

In June 2009, Boyd gave a talk entitled The Not-So-Hidden Politics of Class Online in which she examined the use of two popular social networks, MySpace and Facebook. In particular, she looked at how people have moved from one to the other.

Unsurprisingly, Boyd found this movement to be influenced mainly by non-technical matters, such as whether a person's friends are on there, the social class of other users and the way those users express themselves. Most of the switchers were teenagers. Peer pressure influences everyone but seems always to bear the hardest on the young.

It was a USA-only study, so some of the terminology might be unfamiliar to Britons. I suspect most of the ideas would not be. "White-flight", for instance, describes the way in which a predominantly WASP (white, Anglo-Saxon, Protestant) community withdraws from an area that has a large influx of other, typically darker-skinned people. Here, the area is an online rather than geographical zone.

As Boyd puts it: "Fear of the "other" is core to white flight...". For those who can afford it, that trend ultimately results in the creation of gated communities. In these reverse ghettos, people hide behind physical barriers to avoid contact with people they don't like or whom they fear or both.

Monarchs and princes have done this for centuries, also having their own protected routes between properties to avoid contact with the great unwashed. The Vasari Corridor in Florence is a dramatic example.

These physical safe areas and protected routes were not and are not created by teenagers. Boyd continues: "In many ways, adult worlds are even more divided than teen worlds. Adults are less likely to know other adults who aren't like them than teens are."

Outside the workplace, adults have a freer choice of associates than young people. They have a freer choice of almost everything. However much teenagers might think of themselves as arbiters of what's important - a belief encouraged by marketeers, for less than altruistic reasons - oldies run the world.

Are there any implications here for the business use of social networking? There are three I can think of straightaway:

1. Whatever the technical merits of the software or service you choose, the intended users must want to communicate, to form a community. Without some commonality of purpose, beliefs, background and the like, that community is unlikely to cohere or to last. Look at any failed groupware installation and this lack will often be the cause.

2. Conversely, an online community might need protecting against dilution by 'others'. Allowing in people who don't share the community's values and aspirations could give rise to challenging behaviour (teacher-speak for being awkward), the creation of factions and the loss of critical mass. Conversely, for existing communities, it might reinvigorate what has become a stale, hidebound and inward-looking group. This has to be a matter of judgement of the condition and life-stage of the community.

3. Technical considerations such as cost, ease of use, range of features and ease of administration do make a difference. They modulate the kind and frequency of conversations among users. However, good ratings in each of those categories don't guarantee a lively, constructive community - these depend on the social factors as discussed. On the other hand, a system scoring poorly on the technical aspects makes such a community hard to achieve and may strangle it at birth.

It's worth going to the trouble to select a good system, therefore, but don't expect it to do the job of community creation and sustenance for you..

Social networking for grown-ups

Another study published recently has been a survey of people using Facebook. Again, it looks only at the USA. I was surprised to see that that country accounts for only a third of the site's 200 million users worldwide.

Facebook's demographics and statistics is an interesting antidote to the common perception that this is a medium only or even mainly for teenagers. They barely accounted for a tenth of the total number of users. The largest age group is people aged between 35 and 54, with 1 in 12 users being older than that. It's safe for grown-ups is the message.

Where young people now go to get away from their parents and other wrinklies is the question the authors raise. No doubt someone will soon be along with some ideas and even some numbers.