Free Republic
Browse · Search
News/Activism
Topics · Post Article

Skip to comments.

Numerical Analysis: Newsweek poll - September 2-3, 2004.
Yahoo Financial News: PRNewswire - Newsweek Poll: Republican Convention 2004 ^ | September 5th, 2004 | dvwjr

Posted on 09/05/2004 4:01:55 AM PDT by dvwjr

click here to read article


Navigation: use the links below to view more comments.
first previous 1-2021-4041-6061-8081-87 last
To: dvwjr

Bump


81 posted on 09/05/2004 1:34:12 PM PDT by mattdono ([John Edwards before going on stage 20 minutes after President] Hold muh' senses)
[ Post Reply | Private Reply | To 1 | View Replies]

To: sitetest
"I think this is why pollsters' random samples see increases in the affiliation of the party that just held its convention. It isn't a long-term change, but it does affect the sample in the short term. "

I think there’s not a snow balls chance in hell that the ratio of Republicans to Democrats shifted 20% overnight. I think that anyone who believes that doesn’t know human nature.

82 posted on 09/05/2004 1:44:43 PM PDT by elfman2
[ Post Reply | Private Reply | To 76 | View Replies]

To: elfman2

Dear elfman2,

"I think there’s not a snow balls chance in hell that the ratio of Republicans to Democrats shifted 20% overnight."

I'd probably give it a bit more than a snowball's chance in Hell, but I didn't state that that accounted for the entire effect. That's why I said, "small but significant percentage of respondents."

Here's another reason why random calling might produce more Republicans after the Republican Convention, and more Democrats after the Democrat Convention: effect of the conventions on refusal rates.

I guess that if a pollster considered differing refusal rates as responsible for part of the effect, he'd want to correct for it. And it appears that Newsweek did correct, to a degree, for Republican oversampling.

The point is that there are likely multiple reasons for the oversampling, and not all of it should be corrected.

And I'll reiterate that focusing strictly on the ratio of R/D/I may also be an error to avoid. There are other demographic data for which the pollsters may wish to make some corrections.


sitetest


83 posted on 09/05/2004 2:18:03 PM PDT by sitetest (Spitball Kerry for Collaborator-in-Chief!)
[ Post Reply | Private Reply | To 82 | View Replies]

To: dvwjr
Excellent work. Just plain excellent.

Please keep me on your ping list!

84 posted on 09/05/2004 2:45:01 PM PDT by Dont Mention the War (we use the ¡°ml maximize¡± command in Stata to obtain estimates of each aj , bj, and cm.)
[ Post Reply | Private Reply | To 2 | View Replies]

To: dvwjr

Well .. I see "registered voters" and no LIKELY VOTERS (which are always higher for Bush). I also see more dems than repubs .. along with more WOMEN (who tend to vote dem) than MEN (who tend to vote repub).

So .. I would keep Kerry in the 37-39% range (the Bush-haters group) and lift Bush up to 58-60%.


85 posted on 09/05/2004 4:05:18 PM PDT by CyberAnt (Sen.Miller said, "Bush is a God-fearing man with a good heart and a spine of tempered steel")
[ Post Reply | Private Reply | To 2 | View Replies]

To: dvwjr

Ping me too, please.

Excellent job!


86 posted on 09/05/2004 6:02:12 PM PDT by BlueCat
[ Post Reply | Private Reply | To 2 | View Replies]

To: dvwjr

I like that 9% of Independents favor Nader. If they refuse to vote for Bush, I am happy they are going to Nader instead of Kerry. My daughter has convinced two non-Bush supporters to go to Nader instead of Kerry.

If anyone is anti-war, it should be easy to convince them to switch over to Nader.


87 posted on 09/07/2004 9:26:40 AM PDT by BushisTheMan
[ Post Reply | Private Reply | To 2 | View Replies]


Navigation: use the links below to view more comments.
first previous 1-2021-4041-6061-8081-87 last

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.

Free Republic
Browse · Search
News/Activism
Topics · Post Article

FreeRepublic, LLC, PO BOX 9771, FRESNO, CA 93794
FreeRepublic.com is powered by software copyright 2000-2008 John Robinson