Daily Kos: Time Poll was Massaged to produce lead

[liveoaktx]

http://www.dailykos.com/story/2004/9/4/35010/97521

Shows why this poll is not only flawed but unfair.

[chiburb]

1. Here's the salient point:

Snip:

Please Note that this polling company did something unheard of during a convention they pushed the the undecided likely voters to make a choice. Also again I want to point out that this poll was done during the RNC while the Time poll for the DNC was done 5-7 days later, not at the height of the convention.

Note the dates of 8/24-26 in which Kerry and Bush are tied 46-46 just prior to the Republican convention. In the LV section this date has an asterisk next to it.The asterisk says that in that poll they didn't push the leaners. BUT in the convention poll they did push the leaners. This act alone, just changing from not pushing the leaners to pushing the leaners surely accounts for much of the seemingly large difference. This change in methodology raises questions. And why are they not aware of the methodological problems and why don't they tell us.

Suppose the 2004 election for president were being held today, and you had to choose between John Kerry and John Edwards, the Democrats, and George W. Bush and Dick Cheney, the Republicans. For whom would you vote?" If undecided: "Even though you haven't made up your mind yet, do you lean more toward supporting Kerry or Bush?

Bush/
Cheney Kerry/
Edwards Other
(vol.) Unsure
% % % %
Among likely voters:
8/31 - 9/2/04
53 43 1 3
8/24-26/04*
46 46 - 7
8/3-5/04
44 51 1 4
7/20-22/04
46 50 1 3
.

Among registered voters:
8/31 - 9/2/04
50 42 1 7
8/3-5/04
43 51 1 5
7/20-22/04
45 50 1 4

[liveoaktx]

3. Another salient point is that the methodology skews towards males

"In order to ensure that everyone in the household has an equal chance at being
interviewed, we ask to speak with the youngest male, 18 years of age or older. If
no male is available we ask to speak with the oldest female, 18 years of age or
older. Asking this question helps to increase the level of chance and make sure
that we speak with a diverse population."

Plus, since they polled DURING THE CONVENTION, how much more likely was it that they would get a male REPUBLICAN staying home to watch the convention, as opposed to a male DEMOCRAT - and, of course, they did not use the same time frame for Kerry, but polled 7 days AFTER.

[snippy]

5. Screening the respondents by age and gender does not increase the

level of chance or result in speaking to a diverse population. Exactly the opposite is true, it decreases the level of chance and lessens the diversity of respondents. This poll is not of a random sample because of this methodology.

[GiovanniC]

8. Wait a Minute...

No offense to younger people (I'm only 24 myself), but it seems that the younger you are, the less likely you are to vote. I can think of several households within my circle of family and friends where, if they asked to talk to the youngest male over 18, they'd get someone who doesn't even really vote -- even if other members in his household do. That's a piss-poor way to get LIKELY voters, isn't it?

[Warpy]

6. I don't know about you, but when I get pushed

I tend to push back. Even if I might agree with a push poller, I'll tend to reconsider the whole thing once I hang up. I just don't like being pushed; in fact, I resent the hell out of it.

But then, I've always been on the ornery side.

[liberalmuse]

7. The Daily Kos is a must...

if you want the polls broken down and explained in depth and don't want to go through pages and pages of stuff on the polling sites. He gets to the meat. Excellent work!

[Shoeempress]

2. Thank you!. With two other polls released the same day showing a s

statistical dead heat, the media and the DUs are focusing on this BS Poll. Screw the media, they NEED a turbulent race or no one will watch them. DUH! Why is everyone falling for this? Aren't we supposed to be the Smart folks?

[PA Democrat]

4. Wow! If this does not make a case for unfair manipulation

of polls and how they are reported, I don't know what does. It is also interesting that there are THREE other polls taken during almost the same time period which show Bush and Kerry in a statistical tie.

Interesting that the media chooses this one to report. Also interesting was the slip up by CNN's Judy Woodruff, questioning (when she thinks she is off camera) why they are using the Time poll as opposed to another with very different results.

[quaoar]

9. As the old saying goes

There are lies, damned lies and statistics.

[DeepModem Mom]

10. Nominated -- story needs wider coverage, if any "liberal" media...

would dare cover it. Maybe more net media might pick it up. (I saw the "Salon" story about the Bush family friend included in a report on CNN International -- thought it might have been a first.)

[VOX]

11. This makes me feel sick.

I've been up all night, and have no resistance for this kind of big-brother crap.

This step-out-of-line should be screamed from the mountain tops! :grr: