Internet Poll Accuracy

January 9, 2003 9:22 AM

Join the mission to rig this poll by voting for the red bar. (You can vote a second time by deleting your GuardianUnlimited cookies.)
posted by Pretty_Generic (36 comments total)

Lame.
posted by interrobang at 9:27 AM on January 9, 2003

I thought it would foster debate on the accuracy of internet polling.
posted by Pretty_Generic at 9:29 AM on January 9, 2003

I'm sorry.
posted by interrobang at 9:31 AM on January 9, 2003

Mmm, ambiguous.
posted by Pretty_Generic at 9:34 AM on January 9, 2003

Whats the debate ? They aren't accurate.
posted by zeoslap at 9:34 AM on January 9, 2003

They also matter about as much as an internet petition.
posted by interrobang at 9:37 AM on January 9, 2003

Interrobang, are you sorry about calling the post lame, about it being lame, or about PG's thwarted intention?
posted by signal at 9:37 AM on January 9, 2003

83% to 84% while I was spamming it... you can do eeeet!
posted by zekinskia at 9:38 AM on January 9, 2003

don't you think they would find this link and delete all answers that originated from metafilter?
posted by pxe2000 at 9:40 AM on January 9, 2003

Since the publication is a leftwing morceau de merde, what's the point?
posted by ParisParamus at 9:42 AM on January 9, 2003

Sure, but are they any less accurate than offline polls?
posted by ook at 9:43 AM on January 9, 2003

Signal: I was calling Lame on the tradition of the single-link-post.
posted by interrobang at 9:45 AM on January 9, 2003

The MBTA, which runs Boston's public-tranit system, currently has a poll on whether it sucks. They're not even dropping a cookie, so you can vote early and often.
posted by agaffin at 9:48 AM on January 9, 2003

I think newspolls in general are completely pointless and only serve to detract from objectivity in reporting. Leave it to people who are going to conduct surveys in a scientific way.

I also think the Guardian (the news section, not the opinion section) is about as far from a morceau de merde as you can get in modern journalism. Shall we have a poll?
posted by Pretty_Generic at 9:50 AM on January 9, 2003

Internet polls are important inasmuch as they inadvertently disclose either editorial agendas and/or opinion-forming trends. The Onion's STATshots, for instance, are widely regarded as scientifically sound and precise to an immeasurable degree.
posted by 111 at 9:52 AM on January 9, 2003

Once upon a time, in the forest with no end, there lived a cranky old woman with coal-black eyes and the temperament of a rabid coyote.
posted by Witty at 9:53 AM on January 9, 2003

Interronbang: so you're saying, if the post had contained links to other pointless online polls it wouldn't have been lame?
posted by signal at 9:54 AM on January 9, 2003

Signal: I was calling Lame on the tradition of the single-link-post.

Single-link posts are hardly lame. Mefi used to be full of them, and they were good. It's better to judge the quality of the link itself rather than making arbitrary decisions based on format.
posted by frykitty at 10:00 AM on January 9, 2003

PP: If what you say is true then the same goes for 99.99% of your comments.

But to correct you, the Guardian is a liberal paper in the UK & in the UK (like most of the world) 'liberal' tends to refer to the centre of the political spectrum, not the left. Then again, you probably think the Daily Telegraph is written by commies so I'm guessing the tare on your political scale is slightly off the zero-point...

Back on topic...internet polls only poll the opinions of those folks who are on the 'net for starters which is skewing your sample from the off...
posted by i_cola at 10:03 AM on January 9, 2003

Sure, but are they any less accurate than offline polls?
posted by ook at 9:43 AM PST on January 9

Depends on the offline poll, but as a start: Generally, yes. There's little to no attempt to manage the sample that these polls select, the frame (internet users) is not representative of the general population, I can't imagine that most online polls are rigorously checked for bias in wording.
posted by claxton6 at 10:04 AM on January 9, 2003

frykitty: Yeah, I mean what about this one?
posted by Pretty_Generic at 10:04 AM on January 9, 2003

interrobang: Just to be clear: the point of MetaFilter is to find the best and most interesting of the web to share with others

it doesn't say to "develop a thesis, or muti-link argument about something". Also, if ya notice, the sample on the posting page only has one link. :)
posted by blue_beetle at 10:05 AM on January 9, 2003

Yeah, interrobang! If it weren't for single link posts, I'd have no posts at all!! Vive le Single Link Post!!!
posted by jonson at 10:06 AM on January 9, 2003

Signal: I don't want to add to this pointless discussion of online polls, but it would have been a better post if there were links to, say, scholarly articles about the accuracy of them, or links to some other online polls that were obviously either true or false.

A call to spam a poll about polling is quite farkesque. "Come vote for my picture of my baby! Come over here and show you're [sic] support for the best boobies! Spam this poll!"
posted by interrobang at 10:07 AM on January 9, 2003

Whats the debate ? They aren't accurate.

Yes they are. [Now we have a debate]


Even if most people agree that internet polls (or polls in general) aren't accurate, folks still have a remarkable way of pointing to them when they need to back up an assertion. At least, that what 100% of my poll of me and the person in the next cubicle say.
posted by tolkhan at 10:12 AM on January 9, 2003

interrobang - are you suggesting I have a hidden agenda here? Because thats [sic] pretty funny.

This isn't a normal poll. It's a metapoll, if you will.
posted by Pretty_Generic at 10:13 AM on January 9, 2003

It's "pollstmodern", even.
posted by interrobang at 10:14 AM on January 9, 2003

Oh, you were looking for an argument? I'm sorry, this is abuse.
posted by swerdloff at 10:28 AM on January 9, 2003

swerdloff - that's a tagline, there.
posted by Pretty_Generic at 10:32 AM on January 9, 2003

Lisa: "Statistics show that old people drive at least as well as sleep deprived apes"

also

Kent: Mr. Simpson, how do you respond to the charges that petty vandalism such as graffiti is down eighty percent, while heavy sack-beatings are up a shocking nine hundred percent?
Homer: Aw, people can come up with statistics to prove anything, Kent. Forty percent of all people know that.
posted by blue_beetle at 10:36 AM on January 9, 2003

I get what you were trying to do, Pretty_Generic. Questioning the validity of online polls, while giving instructions on how to stuff the ballot box. Oh, the irony. ;)

On Topic: Online polls can be very accurate if the pollsters do the legwork to forestall problems like requiring registration, rounding errors, et cetera. A lot of research companies are using the 'net to perform surveys. On the other hand, your generic online polls can just be interesting content and serve to spark a debate. Slashdot is a perfect example. From their disclaimer:

This whole thing is wildly inaccurate. Rounding errors, ballot stuffers, dynamic IPs, firewalls. If you're using these numbers to do anything important, you're insane.

posted by maniactown at 10:47 AM on January 9, 2003

Life is not a numbers game for me. So why do we seem to count things if there is really no true answer? ?infinity?

Subject matter is not counted in polls, just the #'s of yes's, no's & undecided.

If you're undecided, then why vote? Because as you are clueless to it, it is nothing that you know fully understand the subject matter then. So, what is the point to track it, #'s of nothingness. That to me = Big waste of time, infinity of worthlessness.

What do you need to be an expert in deciding to make up these polls?
Any polls on the subject?
posted by thomcatspike at 10:53 AM on January 9, 2003

...know fully whoops, know or fully....
posted by thomcatspike at 10:56 AM on January 9, 2003

I can't imagine that most online polls are rigorously checked for bias in wording. [claxton6]

That's what I get for omitting the sarcasm tags, I suppose.

Cynical me suspects, though, that most polls, online or off, are not designed to gauge public opinion so much as they are to influence it (or at least to produce convincing-sounding statistics to support an existing agenda).

At one of my prior residences I was invited to participate in an absurd number of telephone polls -- must've gotten on somebody's list somehow -- and the majority of them had indeed been rigorously checked for bias in wording, but not the way you think. Most were thinly-veiled attempts to get positive-sounding quotes to market one product or another; the political polls were generally not much better. In all but three or four out of the fifty or so polls I answered (over a year-long period) it was easy to guess which answers they were hoping for, often after only a few questions.

Even for the few polls that honestly seemed to be not trying to lead me by the nose -- the set of people willing to spend ten minutes answering pesty questions at dinnertime is a pretty darn self-selected, skewed group. (Especially if, as in my case, they tend to call the same people over and over and over.)

Anecdotal evidence is worth what it's worth, sure; and maybe I just fell in with a bad crowd of unscrupulous pollsters. But the whole process certainly upped my granular sodium intake when it comes to statistics. The very sight of a bar graph makes me woozy nowadays.
posted by ook at 11:11 AM on January 9, 2003

[jackass]

Is this thread lame? (select one)

( ) Yes
( ) No

[/jackass]
posted by DakotaPaul at 1:42 PM on January 9, 2003

Internet For Assholes
This week: Voting Fraud for Assholes

This Getting It article traces the poll fraud to the shenanigans surrounding People Magazine's 1998 "Most Beautiful People" poll. Hank the Angry Drunken Dwarf, of Howard Stern fame, started to win and a tradition was born.

As the meaningless dollops of demagoguery fall time and time again to whichever autocratic asshole wields the best vote-busting program, the obvious question arises as to why corporations still cling to this stupid feature.
posted by haqspan at 4:27 PM on January 9, 2003