Blogs Political Influence Questioned: guess how big the Pew sample was?
The wacky folks over at Pew Internet, the same people who constantly bring out bizarre and contradictory surveys on blogging, have found in their latest survey that despite the high-profile cases such as “Rathergate,” the influence of blogs on the American political landscape was mostly circumstantial, and mostly confined to moderating discussions of news events rather than driving them.
Reuters reports that Michael Cornfield, a senior research consultant at Pew saying that the study dispels the notion that blogs are replacing traditional media as the public’s primary source of information at that Bloggers follow buzz as much as they make it.
But just before you take me to task for referring to Pew as whacky, guess how big the sample of blogs was, remembering that there are over 50 million blogs in the blogosphere and at least 10 million in the United States alone.
No, not 10,000….1,000 wouldn’t be close either. 100? close but not cigar. No folks, the sample number of blogs used for the study: less than 40. Indeed the “40” figure included not just blogs, but “forums, newspapers and television”. Potentially the number could have been 20 or even 10 blogs. Now I’m no scientist and I’ve only formally studied statistics briefly, but blind freddy can work out that such as small sample of blogs can never be accurate, in ANY field of measurement. Indeed, you’ve nearly got to ask yourself with joke studies such as this: does Pew have a hidden agenda?
We looked at posts from 40 blogs, not “less than 40.” The unit of analysis was the post or message, not the blog. Our methodology and goals are spelled out in the report, which may be found at our site. We trust your readers will judge the worthiness of our approach and findings for themselves.
Michael Cornfield
Shit Michael, you just don’t get it do you, 10, 20 or even the 40 blogs you cliam DOES NOT MAKE THE POLITICAL BLOGOSPHERE. F*ck, it would be like taking 40 Australias out of the 20 million and saying that your results reflect all Australians, they wouldn’t and neither in 1 million years would a sample of 40 blog reflect the political blogosphere.
To readers, I apologise for the language, but bugger me dead, they just don’t get it. This shit gets run by the mainstream press based on dodgy sample sizes and is reported as fact by some. Tell you what, I’m going to visit Hicksville, USA, and poll the views of 40 Americans and then put out a statement to say that the finding reflects all Americans: the results would read:
Yes Guns, No Gays, Nuke Iraq and burn the Koran and Women in the Kitchen. Would it be fair if I did this? God no!
Looking at the posts of 40 blogs on any given day does not make a measure for anything when given the sheer number of blogs out there.
Duncan, I agree with your findings completely. Thanks for pointing this out to the blogosphere, as you were the first one to do so. I’ve cited you in my own posting called People who underestimate weblogs.