Slog News & Arts

Line Out

Music & Nightlife

« Confession | Another Toilet Cruiser Busted »

Tuesday, September 18, 2007

Re: Those Slog Tracking Poll Results

posted by on September 18 at 9:40 AM

Last week I promised to Slog about a difference I noticed between our September and August presidential race tracking polls. It’s taken me a bit longer than promised, but here goes.

First, the usual caveats. I know these polls are unscientific. I know I’m only getting a sense of the opinions of a self-selected group of people (poll clicker-oners) within a larger self-selected group of people (Slog readers). Still, I’m interested in what the Slogosphere thinks about the Democratic presidential candidates (or what it tells me it thinks about them) and how what it thinks about them changes over time.

So, let’s look at the August poll results, the ones from the poll that included Gore in the list of choices:

In August, as you can see, Gore was getting 45-percent of the Slog reader vote. Now, let’s look at the results from early September when Gore was included:

In early September, Gore received only 39-percent of the vote, down 6 percentage points. And it seems to me there are more comments like this attached to the September poll posts:

Can you guys get over Gore running? You might as well list Santa Claus and Ghandi in the poll along side Gore.

Posted by John | September 10, 2007 10:57 AM

Is the Gore bubble bursting among Slog readers? Are people over the hope that he’ll swoop into the race at the last minute and provide another alternative to Clinton, Edwards, and Obama (and Dodd, Gravel, Kucinich, and Richardson)? Is this due to a growing satisfaction with the current slate of Democratic candidates or just the end of a flirtation with an interesting rescue narrative?

Or, is the Slog tracking poll’s margin of error much, much greater than +/- 6 percentage points? (Paging all math and statistics brainiacs.)

Something to think about as we wait for next month’s tracking poll. Meanwhile, some additional analysis of September vs. October from our commenters.

RSS icon Comments

1

Or maybe it's the fact that Gore keeps saying, "I WILL NOT RUN."

Posted by Ivan Cockrum | September 18, 2007 9:50 AM
2

Computing the margin of error on this type of poll would be like computing the margin of error on reading chicken entrails.

THAT'S WHAT UNSCIENTIFIC MEANS, YOU MORON.

Posted by elenchos | September 18, 2007 9:50 AM
3

What 2 said (minus "you moron"). This poll is, as they say, not even wrong.

Posted by Levislade | September 18, 2007 9:51 AM
4

I like to see that at least not everyone commenting on slog is totally racist and Obama gained some points. Is he the new model minority?

Posted by hunh? | September 18, 2007 9:55 AM
5

The next president should make Gore Secretary of the Interior. But get over Gore being president himself.

Posted by monkey | September 18, 2007 9:55 AM
6

and how many angels can dance on the head of a pin? maybe the next unscientific poll can be devoted to that burning question.

Posted by ellarosa | September 18, 2007 9:56 AM
7

@4

Uh huh. Anybody who didn't vote for Obama is a racist. Is that you, Ecce Homo? Back from the dead?

Posted by Mr. Poe | September 18, 2007 9:57 AM
8

I was unfortunate enough to be drinking some coffee when I read about this poll's "margin of error" and I spit it up all over my keyboard -- thanks a lot!

Posted by twee | September 18, 2007 10:01 AM
9

Okay Eli, cards on the table. Is it something you're thinking about? Do you think Gore will jump into the race?

Posted by E. Steven | September 18, 2007 10:08 AM
10

Oh I would so vote for Santa Claus.

Posted by Jordyn | September 18, 2007 10:08 AM
11

Are people over Gore? Yes, about 8 years ago.

Is this due to growing satisfaction with current candidates? No.

Is the margin of error much greater than +/-6%? Certainly yes. First, sample size is significantly different (998 in August vs 620 in September). That alone will skew your results.

Who responded in the first poll because seeing Gore listed was funny? It wasn't so funny the second time, so they didn't bother to respond. In August, the "with Gore" poll had 86% the number of respondents as the "without Gore" poll. In September, that dropped to 81%, a difference of 5%!

As for the 39% who still picked Gore in the last poll? I'm guessing they also watch Seinfeld reruns and think they are every bit as funny as when they first aired.

Posted by Mahtli69 | September 18, 2007 10:10 AM
12

@11 - Not that there's anything wrong with that . . .

Posted by Ziggity | September 18, 2007 10:12 AM
13

Ivan Cockrum has it wrong.

Gore says, basically, "I am not running", a statement very different from "will not run".

Posted by N in Seattle | September 18, 2007 10:12 AM
14

I think as the first primary (in whichever state it ends up being, which is clearly not going to be Washington) gets closer, it just looks less and less likely that Gore is going to jump in.

So, the erstwhile Gore people are deciding which of their second choices are most likely to win and/or least objectionable.

Posted by ertyt | September 18, 2007 10:14 AM
15

Oh, Eli, if you're going to ask for advice on "margin of error", it might be nice to tell us the vote count. If we assume that the poll is valid (OK, let's pretend...), it remains impossible to estimate MOE without N.

Posted by N in Seattle | September 18, 2007 10:17 AM
16

@11 Nice analysis, except for the jibe at Seinfeld. Next you're going to tell us that you don't like Chris Crocker. Puhleeze! The vox populi has spoken!! :P

Posted by Katelyn | September 18, 2007 10:24 AM
17

I'm still missing the Obama fascination. The guy's running on personality, not policy. Where's the beef, yo?

Posted by Kim | September 18, 2007 10:33 AM
18

@11 and @15

Not even close. How many votes you counted in an unscientific poll doesn't matter in the least.

It is fun though. The same people who would mock snake handlers and Bush's faith based governing and other superstitious fools will smugly discuss a nonsense poll like this all day. I really enjoy seeing someone scoff at those who "wasted" their vote on Gore in this pseudo poll. Do you have any idea how ironic that is?

I'm waiting for someone to ask whether or not this poll has electrolytes.

Posted by elenchos | September 18, 2007 10:45 AM
19

I think in the movie it was made clear that toilet water does NOT have electrolytes, elenchos. So that should answer the question.

Posted by Katelyn | September 18, 2007 10:52 AM
20

Gore should not run. When he lost the presidency, he had the stink of loser on him. Conceding the presidency and then fighting it in court and such... it was a big, long drawn out loss. He recovered and right now, he's the man. People love him for what he is. If he ran again and lost once more, he'd have the stink of mega-loser on him. He should just keep on doing what he does and continue being the man, the man the people love. His life is really good right now. He's doing his thing, hanging out with celebrities, winning all kinds of awards and get accolades. For strictly selfish reasons, why fuck with that?

Posted by JC | September 18, 2007 10:54 AM
21

@18 - elenchos, 1 out of 1 person polled think you are a buffoon. That's 100%! The number of people counted doesn't matter in the least.

Posted by Mahtli69 | September 18, 2007 11:08 AM
22

The difference in sample size only changes the size of the random error. But the problem with a self-selected sample isn't the random error. It's the systemic bias caused by whatever motivates people to vote or not vote in the poll.

Look. If you think the sample size "skews" the result, then answer a simple question. Which direction is the result skewed? Does the larger sample favor Gore? Or does it favor someone else? Most importantly, how do you know?

I might be a buffoon, but I can stop being a buffoon any time I please. You, however, can never change the fact that you were wrong about this. And very likely, you will always be wrong about it, because you don't want to learn.

Posted by elenchos | September 18, 2007 2:56 PM
23

I think it's just that a lot of us Gore/Obama people are willing to vote Obama/Gore if need be.

Posted by Will in Seattle | September 18, 2007 3:21 PM
24

@22 - The difference in sample size only changes the size of the random error in a scientific survey, which everyone agrees this is not.

My position is the sample size difference was not a random error, and more participation in the first survey skewed the vote towards Gore for the reasons I explained in @11 and here (@6).

There were less participants in the 2nd poll and, of the people who did participate, a smaller percentage bothered to fill out the 2nd part which included Gore.

The following is only my opinion and is definitely not scientific. I have no way to confirm any of this (other than taking a poll):
1. The "wow factor" of taking part in a poll with Gore as a candidate was significantly decreased the 2nd time around. Accordingly, Gore supporters were less likely to take part.
2. For the same reason, the 5% drop in "with Gore" participation between the two surveys was made up of a larger than typical proportion of Gore voters.
3. Calling you a buffoon was not necessary. Sorry.

I'll add one more thing ... If Gore was actually running, I don't think he'd be polling 45%, or even 39%. He'd be in the pack with Clinton, Obama, and Edwards. The fact that he's not running means we can all idealize him (just like Santa Claus and Gandhi).

Posted by Mahtli69 | September 18, 2007 4:33 PM
25

Look Al-Gore-Rhythms: he's won an Oscar, now an Emmy. Why should he run for president when he has the oppportunity to join the very select group of nine (who've won the Big Four awards) by winning a Grammy ("Al Gore Speaks") and a Tony ("Internezzo: the Musical")? Then he could become the fourth elected president to win a Nobel. Google aweigh, kids.

Posted by KY. COL. of TRUTH | September 19, 2007 12:03 PM
26

pardon the interuption kiddies, but i just found some video on google, that, i think, should be addressed by presidential candidates.
norman minetta's testimony infront of congress http://video.google.com/videoplay?docid=-3722436852417384871

and a big explosion sound
http://video.google.com/videoplay?docid=-5182535448932065917

Posted by fremont grandma | September 19, 2007 12:18 PM

Comments Closed

In order to combat spam, we are no longer accepting comments on this post (or any post more than 14 days old).