Study revives six degrees theory

The article continues at http://news.bbc.co.uk/2/hi/technology/7539329.stm

Very interesting results.

The article seems to be tracking instant messages. Is the article ignoring people who don't have internet access?

Seeing as the study was done by Microsoft researchers using their own company's databases.

Consider telephone polls. By definition, such polls include only those people who A) have phones, B) answer the phone and C) are willing to answer all the damned questions. However, this sample represents a broad enough cross-section of American that a very confident extrapolation of this group to Americans as a whole can be made.

I bet that a similar high confidence is found when comparing those who use Microsoft IM and internet users in general, enough that a sample size of the IM users can be extrapolated to all internet users. Given the size of the sample -- looking at everyone who received or sent a text message over their system in a one month time -- their results probably have a lower margin of error that most phone polls.