The Harvard College Democrats
(shield)
(shield)

17739
DOORS

1732
CALLS

User login

Syndicate

Syndicate content

A quick object lesson in poll analysis

Posted on Thu, 03/13/2008 - 1:09am by Markus Kolic

Red flag #1 ought to have been that this was in the Washington Times, a paper that's not exactly known for its journalistic rigor. But no, I soldiered on, reading a fairly bland writeup about a Harris poll that shows -- Surprise! -- not everybody reads political blogs. Whatever. Until this:

The Harris poll, meanwhile, found that political blog readership was lowest among those younger than 40 — and highest among people 63 and older.

Buh? Since when are seniors the biggest readers of political blogs? I mean, I know I do my best to court that demographic, but I'm fairly sure I'm in the minority there. One does not imagine Grampa gettin' up in the morning and checking out, for example, TBogg, especially not more frequently than the average young person. So -- all you amateur statisticians out there, can you think of a reason this finding might be flawed? How about we read the next sentence?

The online Harris survey of 2,302 adults was conducted Jan. 15-22.

Aha! The only people who could respond to this survey -- in polling terminology, its "sample frame" -- were people who were already using the Internet. Now, we can assume that it's a fairly small and non-representative portion of seniors who are online regularly, whereas a larger and more diverse (though still non-representative) portion of younger people use the Internet. (See Gallup, among many other sources.) So you cannot generalize from these findings out to the population at large, only to Internet users, and hence saying that blog readership is "higher" among one group or another is simply wrong. In fact, it's likely that in raw numbers, more young and middle-aged people read political blogs than older people, simply because there's more of them online. This is very very simple stuff, and it's sad that no editor at the Washington Times (assuming they have editors) caught it.

(This is also why Internet polls -- like the Harris and Zogby Interactive -- are completely worthless for political purposes. The Internet just cannot provide a representative sample. Or at least not yet; if current trends continue, Internet use may be as standard as phone use within a decade or so, at which point it'll be kosher for polling.)

...Still, maybe I'm wrong, and the chief demographic of a political blog really is the over-65 set. In which case, what the hell, give 'em what they want has always been my motto...


----SHAMELESS CROSS-PROMOTION ALERT: Grab the Indy today (or, if you're over 65, read it online) for a bunch of insightful pieces on the 5th anniversary of Iraq, including a brilliant analysis of media's role in war and public discourse by our own Sam Jack.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Thanks for the plug! Markus

Posted on Thu, 03/13/2008 - 2:17am by Sam Jack

Thanks for the plug! Markus is too modest to mention that his own piece is also in the issue, and a smashing piece it is.

And I have to agree with you on the blogging thing. My own grandmother is a blog reader (of this blog, incidentally, hi grandma!), but I wouldn't call her an average exemplar of the 65 and older crowd. She actually can use the computer--for a variety of purposes. My grandfather on the other side of the family, on the other hand, I don't think he even knows how to use a mouse.

Grandma was an early adopter; I remember playing with an old Macintosh 128k that she got in 1984. So if you wanted to capture the subset of seniors who are also tech junkies, an online poll is the way to go.