The latest Pew study on blogging seems to be turning up everywhere, and thanks to LiB I even took the online followup survey.
And I’m even more suspicious than usual about one reported figure–noting that, with a sample size of 232 people willing to engage in a lengthy telephone survey, I’d treat all of the numbers as vague guesstimates anyway.
The really suspicious number: That only 15% of bloggers provide feeds.
Which, in the online survey at least, is specifically worded as “RSS feed.”
Want to bet that some significant percentage of bloggers don’t turn off the default feeds that are set up by most blogging software, but aren’t aware that they’ve enabled something called “RSS”?
Want to bet that loads of Bloglines users don’t know they’re doing “RSS” when they subscribe to blog? Particularly a blog with an Atom feed? (Or a Feedburner feed, or…)
If the question was worded “Can people subscribe to your blog?” it might get different answers. Or it might not: I’d guess many bloggers don’t even know their blog is available via aggregators.
[Then there are questions that involve navel-gazing beyond my level, e.g., How many other blogs link to mine? I put "15," since "I have no idea, but some do" is a bad answer. Also, for those of us, ahem, who don't have blogrolls but do link to our public Bloglines lists, the question as to how many blogs we link to is a little tricky...]
Anyway: Interesting stuff. By Pew’s own admission, this one requires considerably more than the usual grain of salt that most surveys require: Remember that 10% of a 232-person sample is 23 people, no matter how much fun it is to scale it to 30 million Americans. And no matter how many newspapers report it in the latter form.
(For those questions with 2,000-person responses, if the question’s a yes/no question, there’s some plausible reason to believe you can project the answer. Maybe. If a whole bunch of other assumptions are correct–and, statistical methods aside, these are “social science” issues more than they are mathematical issues. For questions with 232-person responses and multiple answers…well, they’re interesting responses. )