Archived material Access restricted Article
Rantburg

Today's Front Page   View All of Sat 10/02/2004 View Fri 10/01/2004 View Thu 09/30/2004 View Wed 09/29/2004 View Tue 09/28/2004 View Mon 09/27/2004 View Sun 09/26/2004
1
2004-10-02 Home Front: Politix
For crying out loud, Newsweek shows Kerry with a 2 point lead...
Archived material is restricted to Rantburg regulars and members. If you need access email fred.pruitt=at=gmail.com with your nick to be added to the members list. There is no charge to join Rantburg as a member.
Posted by RJB in JC MO 2004-10-02 7:26:03 PM|| || Front Page|| [3 views since 2007-05-07]  Top

#1 A few problesm with this.

1) It was a "push" poll - with questions designed to lead people into a Kerry selectiona t the end of all questions. (Note the topics, expecially the draft question).

2) They sample and thne "correct" the sampel to match exit polls from 2000, which means they adjust it to have more democrats and a larger independant fraction.

3) Its registered voters, not likely voters.

4) Their results are completely different from 5 other polls done in the same time frame.

Basically, this is about as real as the CBS story on the draft, and the 60 Minutes story on Bush's guard duty.
Posted by OldSpook 2004-10-02 8:09:00 PM||   2004-10-02 8:09:00 PM|| Front Page Top

#2 Just take this one point:

Kerry, typically characterized as aloof and out of touch by his opponents, came across as more personally likeable than Bush (47 percent to the president’s 41 percent).

This has NEVER been above 40% for Kerry. ANd the 4 other surveys, including one by Carvile's organization (overnight after the debate) consistently show Bush above 50% on this number, and Kerry below 40%, even after the debate.

Newsweek is lying or else distorting their numbers.
Posted by OldSpook 2004-10-02 8:12:37 PM||   2004-10-02 8:12:37 PM|| Front Page Top

#3 For example, look at the professionals. Rasumussen, who is percieved to have a 3% ro so bias toward Democrats in his methodology, still has Bush up by more than the margin of error. Sampled over the same time the Newsweek Poll was done. Bush has 49% rounded down from 49.4, Kerry has 46% rounded up from 45.6 with less than 3% margin of error.

(from Rasumussen)
emphasis mine
These results are based upon a survey of 3,000 Likely Voters conducted Wednesday, Thursday, and Friday night. As a result, just over one-third of the interviews were conducted following Thursday night’s Presidential Debate. Results from last night’s sample were similar to the preceding nights
Posted by OldSpook 2004-10-02 8:18:33 PM||   2004-10-02 8:18:33 PM|| Front Page Top

#4 Truman shoved the pollsters numbers and the newspaper headlines up their asses in his election. It is extremely difficult to get a good poll number. You first have to define your representative sample, that is a major consideration. Then you have to frame the question correctly. Blah blah. Then you have do deal with the objectivity of the pollster. More blah blah.

Bottom line: the only poll that will count is the November voting poll.

BTW, do you notice that there is always a passive agressive opposition by the established parties to purging, cross checking, and updating voter rolls?
Posted by Alaska Paul 2004-10-02 8:23:53 PM||   2004-10-02 8:23:53 PM|| Front Page Top

#5 Finally dug out the sample size:

36Dem
34Rep
30 Indpt/Other

Far overweighting Democrats given the change in Party Identification that all other polls have had. Most other national polls have the reverse percentages in terms of PID.

So, yes, if you overwieght Democrats, and call other democrats "independants", then of course you can produce a pol that shows Kerry in the lead.

Happened all the time in the California poll that showed Schwarzenaegger losing, or the Colorado senate race between that had similar sampling mistakes to produce the Demcorat with a 5% lead, who then lost by 6% to the Republican.
Posted by OldSpook 2004-10-02 8:24:03 PM||   2004-10-02 8:24:03 PM|| Front Page Top

#6 OldSpook---That's OK. The more that these pollsters screw the statistical pooch, the less credibility they have. Just like CBS and Rather in the Bush national guard non-story.

The thing that is bad in elections is like what happened in Florida. The errors detected in the balloting were within the noise band created by errors and possible fraud in the voter rolls, so how you decide to handle the errors or ballot problems will affect the outcome of the election. It is veddy veddy bad to make decisions in the noise band without appropriate noise filter algorithms. **wags finger**
Posted by Alaska Paul 2004-10-02 8:32:32 PM||   2004-10-02 8:32:32 PM|| Front Page Top

#7 All true, but the fact remains that Kerry's supporters were losing heart, and now they've got some hope again. Bush and Rove need to slam Kerry, and hard, on his global test and EU help-on-the-way idiocies. And slam him twice as hard on his freezenik proposal to unilaterally disarm us by shutting down bunker-buster production. This is crunch time.
Posted by lex 2004-10-02 8:40:38 PM||   2004-10-02 8:40:38 PM|| Front Page Top

#8 Want to break their hearts cancel the rest of the debates.
Posted by Sock Puppet of Doom  2004-10-02 8:43:14 PM|| [http://www.slhess.com]  2004-10-02 8:43:14 PM|| Front Page Top

#9 Bush never should have agreed to three debates. He and Rove need to go into overdrive now with negative ads slamming Kerry as an unreconstructed eighties-era peacenik/freexe advocate/idiot. Bunker-busters should be Topic A.
Posted by lex 2004-10-02 8:51:12 PM||   2004-10-02 8:51:12 PM|| Front Page Top

#10 Oh, and ratchet up the offensives in Iraq. Bring it to a conclusion. Force the MSM to either support the troops or try to pull another Cronkite, which would backfire big time. Their credibiity's shot. Sumner Redstone will not permit more tricks from Rather and Co.
Posted by lex 2004-10-02 8:52:55 PM||   2004-10-02 8:52:55 PM|| Front Page Top

#11 I do not remember a debate during wartime on the war itself. Is this a first?
Posted by JP 2004-10-02 11:23:03 PM||   2004-10-02 11:23:03 PM|| Front Page Top

07:03 indu ju killer
07:22 indu ju killer
07:15 indu ju killer
02:55 Bryan
00:26 Bryan
00:05 RWV
23:56 lex
23:52 lex
23:45 Bryan
23:37 JP
23:31 Bryan
23:25 True German Ally
23:23 JP
23:21 True German Ally
23:18 RJB in JC MO
23:15 trailing wife
23:11 lex
23:11 eLarson
23:08 True German Ally
23:05 True German Ally
23:01 Bryan
22:54 jawa
22:52 jawa
22:50 Bryan









Paypal:
Google
Search WWW Search rantburg.com