In message <[EMAIL PROTECTED]>, Rob Nelson <[EMAIL PROTECTED]> wrote: > I dunno, I can't take a survey of 347 users as anything approaching an > accurate survey. That's less than .0000001% of the population of > America.
Actually, .0000001% of the US population would be less than one person. According to the article cited, the poll has a margin of error of 5.4%; so we can expect a 95% chance that, for example, between 32.6% and 43.4% of the population the sample was selected from use unencrypted wireless networks. It is a statistically valid sample size. In any case, consumer host security is a very real problem. -- Shields.
