« Another Rathergate Blog | Main | R.I.P. Slutwear »

Pop Quiz

This is a test to see which party you are in:

The Pew poll found the race at 46-46 among registered voters, and 47-46 Bush among likely voters. A Gallup poll being released Friday has Bush up 54-40 in a three-way matchup, with Ralph Nader at 3 percent.

Quick, which poll is right?

In one they are tied and in one Bush is up 14 points? The poll you believe is probably more indicative of which guy you are voting for than the accuracy of each poll. Having said that, (being a Bush guy) I think it is closer to 10 points right now.

Bush had multiple polls that put him up by double digits after the convention. Kerry, thanks to Dan Rather, has not had a single day to be on message. As such, I find it impossible that a guy who is not doing anything can make up 14 points in a week.

UPDATE: The Pew people are whacked. They broke of the poll into 2 waves. Here's how they describe it:

The latest national survey of 1,972 registered voters by the Pew Research Center for the People & the Press, conducted in two waves over a seven-day period, finds that the president's large margin of support in the initial period (Sept. 8-10) dissipated in the polling conducted Sept. 11-14. Among all registered voters Bush initially led Kerry by 52%-40%. However, the second wave of interviewing shows the race even among registered voters, at 46%-46%

We are to believe the 9/11 anniversary gave Kerry a 12 point bounce over night? Excuse me if I don't buy that one.


TrackBack

Listed below are links to weblogs that reference Pop Quiz:

» Sharp as a Marble linked with Anecdotes to Live By

Comments (17)

O. K.; I'm yer' hucklebe... (Below threshold)

O. K.; I'm yer' huckleberry!


Poles don't really matter in our election process; they can only vote in Poland.


Signed,


Not A. Troll

I don't see Rasmussen (sp?)... (Below threshold)
-S-:

I don't see Rasmussen (sp?) anywhere involved, but based upon what I've read and heard about Rasmussen's polling process, that one's the most reliable.

Maybe the question of THIS polling process should be, rather, which party are Gallup and Pew aligned with, as organizations? 'Cause, EVERYbody has an opinion, even pollsters, even polling organizations. And, polls are often used to sway public opinion, err, something like that (d'oh).

While I'm no huckleberry, I am a Blueberry and I think that information flows far faster than pollsters can readily integrate in their processes, such that what you see reflected in polls are the opinions from a while ago, mostly. Or, else, opinions of people who actually like telemarketing calls at dinner -- and respond to them. Makes you wonder, now, doesn't it, who is included in polls such as Gallup that rely on telephoning home numbers without invite.

Bush is certainly ahead in the process, in my estimation, based upon the grumblings and sentiments of most Americans that I am aware of. Kerry's long since lost hope of favor with the majority, not that he ever had that.

Well, my inclination is to ... (Below threshold)
Chris James:

Well, my inclination is to beleive the Pew poll showing them neck and neck. I'm a registered Independent, and a strong Bush supporter (including financially).

Basically, I *want* to beleive Bush is ahead, but I don't trust polls. I guess I'm a pessimist because I did beleive polls when they said Kerry was ahead.

I also don't want to see anyone getting over-confident; it's still six weeks until election day, and that's the only poll that counts.

I remember reading that on the eve of the 1980 election, all the pollsters were calling it a dead heat...

AAck! BLACKBERRY not BLUEB... (Below threshold)
-S-:

AAck! BLACKBERRY not BLUEBERRY!

Must have caffeine...

1) You didn't say what the ... (Below threshold)
AEN:

1) You didn't say what the group was on the Gallup poll.
2) Has anyone ever done a re-poll of "likely voters" after the election to see what percent of them actually did vote.
3) From what I've seen, Kerry has been ignoring the whole Rather flap.

One more thought; It... (Below threshold)
Chris James:

One more thought;
It's true that Rathergate has kept Kerry out of the spotlight for a while, but why assume that is bad for Kerry? He's got a strong unfavorable rating with voters, and has tended to go up in the polls when he's out of the public view.

Kerry's best strategy at this point might be to take a two-month vacation from campaigning, starting now.

I work for a market researc... (Below threshold)

I work for a market research firm, am a statistician and survey expert, a writer and consultant, among other things. I can tell you that there is more nonsense and garbage going on with these polls than anything I have seen in my long career.

Things you absolutely have to know to gauge the accuracy of these polls:

1) What was the sample size?
2) How was the sample stratified?
3) Did the polling organization use RDD (randomly dialed digits) in their survey efforts?
4) Did they use a panel (previously surveyed on the same subject by the same polling organization)? [Note that this ties into item #3 above.]
5) How were the questions worded?
6) What was the Don't Know/Refused/Uncertain percentage?
7) What was the population base upon which the survey was sampled (likely voters, registered voters, etc.)?
8) What did they use for their prospective contact base? Voter rolls? Screener questions calling homes/numbers randomly?
9) What are the political leanings of the polling organization? If anyone thinks Zogby is unbiased, they are out of their minds.
10) Was Paul Krugman involved in any way in the polling effort?

I could go on, but hey, enough's enough. I love this President and he is going to win.

Wizbang! is amazing.


"based upon what I've read ... (Below threshold)
Jumbo:

"based upon what I've read and heard about Rasmussen's polling process, that one's the most reliable."

I don't know -S-, the Powerline guys (forgive me, Kevin!) lump them with the Star Tribune as far as skewed polling. And as for Pew's, shall we say, new and exciting methods, it sounds like they're eager to keep it close. Hmmm; people don't need to hire pollers when the game is over. So we may see interesting poll demographics in order to "keep it fair": "Bag ladies favor Kerry 4-to-1!"; "Death Row inmates huge for Kerry!", etc.

arrrrgggg I forgot the link... (Below threshold)
Paul:

arrrrgggg I forgot the links... must sleep.

Alright, Jumbo, I'll take a... (Below threshold)
-S-:

Alright, Jumbo, I'll take a finer look...

My point, overview here, is that I tend to read polls often but I also read them within context of who created the polls, who was polled, when and how.

Zogby...I used to consider Zogby reliable but lately, don't know.

Based upon how I've seen the instant polling process so often manipulated (mostly by the Left based upon social memes as they develope and are promoted), I try to stay out of relying on any polling process to exemplify much of anything real or certain.

And rely on other characteristics. But, I DO read POLLING PLACE, that offers a variety of information, and try to exclude most any process that poses questions so skewered as to create their own answers.

I'm going to vote for George Bush, lest anyone wonder, or, not. Just being clear. But I will take another look at Rasmussen, not that huge issues hinge on finding or disproving credibility there. Thanks for the information, Jumbo.

Argh, I just woke up, I nee... (Below threshold)
-S-:

Argh, I just woke up, I need to have some coffee...I'm making so many errors here, sorry.

POLLING REPORT, not "POLLING PLACE."

Sorry, I'm gonna exit until I wake up.

I noticed that this poll wh... (Below threshold)

I noticed that this poll which shows Bush and Kerry even has been all over the airwaves -- TV and radio. But I tend to favor Gallup as non-partisan. Pew and Zogby have always tended to favor the Dems. The last LA Times poll showed Kerry ahead but it was discovered they interviewed 20 percent more Democrats than Republicans. When Sean Hannity questioned the pollster, the reply was, "It doesn't matter." Hmmmm. She must have attended the Daniel Rather School of Journalism.

KCTrio,Good post a... (Below threshold)
Greg F:

KCTrio,

Good post and I agree. The main difference between the polls is how they adjust for party affiliation. In a truly random sample the party affiliation percentages should be close to the party affiliation percentages of the population. The pollsters assume they do not have a random sample and adjust to what they think the party affiliation percentages should be. Rasmussen assumes a higher percentage of Democrats relative to Republicans then do many of the other polls, thus, he finds a closer race. This adjustment is statistical nonsense. It is my view the polls will be off considerably in this election. The country is at war, this is serious business. People tend to put off, or oscillate, on decisions where there are grave consequences, this is impossible to poll beforehand. For many, this time of reckoning will happen when he or she steps into the voting booth. IMO this will be a one issue election, the war on terror.

Greg F et al:Thank... (Below threshold)

Greg F et al:

Thanks for your kind input. Here's some more detail on my point regarding these polls. When one polls any population, one must know the demographic make-up of that population. In the case of consumer polling, it comes down to age and income. These are the only 2 variables that count in ANY consumer poll regarding preferences for coke, the color red or who's your favorite actor.

When dealing with political polling, you still must go through the same process. You start with some basic assumptions about who is going to vote in a given election, then you stratify that population by age and income. You should have a good idea of the age and income level of those that are probably (likely) going to vote in a given election. This is probably based upon historical voting. Some states allow party organizations to obtain registerd voter lists by party affiliation; others only give the percentage of voters in a precinct that voted one way or the other (Tennessee does it this way, BTW).

Now, even knowing the registered voters in a given population (Florida, or the US or a county or MSA) doesn't get at the fundamental questions: 1) Are you going to vote; and, 2) Who are you going to vote for.

The only way to do this accurately is to stratify the population of potential voters, screen them the way I said before, get them to answer if they will vote and for whom, then get their age and income. Now that you've done that, you try to get your responses to match the age and income strata of the voting population.

The next step is to WEIGHT your responses for your strata (the % of the population that falls into each age and income bracket). You NEVER get the completed sample (surveys) to match the population strata, so you must apply a multiplier effect to each survey to get an accurate reflection of how the true vote will turn out if the people that answer your questions vote the way they say they will.

Without the weighting, the results are MEANINGLESS! Say you get 30% of your respondents as under age 30 with household income below $50K. But suppose the voting population has only 10% in this category. Without weighting, your results, if published raw, will be totallly skewed. Same could be said for if you get more seniors, or more people that own their own home, etc.

Of course, age/income strata aren't the ONLY way to do consumer polling; some political pollsters may have to adjust for registered voters. But that gets dicey, because just because someone's registered doesn't mean that they'll consistantly vote that way.

This is the problem with Internet polling. I don't care what anyone says, the Internet public does not yet represent the general population. There is no way to adjust an Internet poll to reflect the general population at large (though there are those out there that would love to say that I'm insane and that they're right).

Another problem with political polls is the use of panels. You call the same people week-to-week and ask them the same sets of questions. These panels make the cost of polling much cheaper, but they ruin your results...the act of polling the same people creates a biased set of poll takers. They've become conditioned to taking polls. Big problem.

This is why Roper Starch (now NOP World) is the premier polling organization in the world. They don't do anything but the right way. No panels, no clients writing the questions the way they want. They own the entire process, and won't even allow a client to spin the results (they retain final right of refusal before their name is associated with a press release, ad or any other such thing resulting from a study they've conducted).

I think Rasmussen seems to be the best out there these days. I highly recommend their premium service, by the way. It's worth every penny.

Again, keep Paul Krugman away from any polling you may do.

Bush will win. All of the signs are in place. All Bush has to do to seel the deal is to simply be himself in the debates and be honest. Kerry is a master orator when it comes to transporting across the room on an issue. Bush needs to keep it simple and honest. Don't get caught up in any policy discussions. Just hammer him as a flip flopper and an anarchist/Chomsky devotee/insane communist (OK, maybe that won't fly, but it's true, nonetheless).

Best regards to all. Sorry for the long post.

Nicely done, all. -S-, you'... (Below threshold)
Jumbo:

Nicely done, all. -S-, you're far more qualified to judge pollers than I' I'm one of those "anecdotalists"! And I tend to agree with you re Zogby but again, only by "feel".

Generally, I (sorry) feel most "civilians" (non-demographers) tend to look at poll numbers and judge them by what the person thinks; sort of a "that sounds about right" test. And KCTrio, I agree it's Bush's to lose. I don't think Kerry can do do anything to win it, but an absolute shocking Iraq disaster or some inarguably proven and relevant lie/deceit by Bush can lose it for Bush.

First, the tied poll uses r... (Below threshold)
David Addams:

First, the tied poll uses registered voters instead of likely voters. This adds numbers for Kerry when compared to other polls, at least the ones using likely voters.
Second, the tied poll was done,in part, over a week-end. Week-end polls usually wind-up over sampling Democratic voters.

First, the tied poll uses r... (Below threshold)
David Addams:

First, the tied poll uses registered voters instead of likely voters. This adds numbers for Kerry when compared to other polls, at least the ones using likely voters.
Second, the tied poll was done,in part, over a week-end. Week-end polls usually wind-up over sampling Democratic voters.




Advertisements









rightads.gif

beltwaybloggers.gif

insiderslogo.jpg

mba_blue.gif

Follow Wizbang

Follow Wizbang on FacebookFollow Wizbang on TwitterSubscribe to Wizbang feedWizbang Mobile

Contact

Send e-mail tips to us:

[email protected]

Fresh Links

Credits

Section Editor: Maggie Whitton

Editors: Jay Tea, Lorie Byrd, Kim Priestap, DJ Drummond, Michael Laprarie, Baron Von Ottomatic, Shawn Mallow, Rick, Dan Karipides, Michael Avitablile, Charlie Quidnunc, Steve Schippert

Emeritus: Paul, Mary Katherine Ham, Jim Addison, Alexander K. McClure, Cassy Fiano, Bill Jempty, John Stansbury, Rob Port

In Memorium: HughS

All original content copyright © 2003-2010 by Wizbang®, LLC. All rights reserved. Wizbang® is a registered service mark.

Powered by Movable Type Pro 4.361

Hosting by ServInt

Ratings on this site are powered by the Ajax Ratings Pro plugin for Movable Type.

Search on this site is powered by the FastSearch plugin for Movable Type.

Blogrolls on this site are powered by the MT-Blogroll.

Temporary site design is based on Cutline and Cutline for MT. Graphics by Apothegm Designs.

Author Login



Terms Of Service

DCMA Compliance Notice

Privacy Policy