According to polls Clinton winning/Obama winning;...Clinton sunk/Obama sunk

If you read the threads on DU, you are likely to start feeling schizoid. In one thread Obama has widened his lead, while in another Clinton is surging forward.

Maybe we should give the polls a rest, at least until Pennsylvania gets closer. I'd say until November.

The obsession with the daily horse-race clouds the mind, and distracts from any valid comparison of the candidates. It has more to do with day-to-day fluctuations and the mindless messages from the media than anything else.

beating or losing to McCain.

Guess most of us have a little political junkie inside :hi:

It is confusing. Polls at this stage are probably meaningless for reflecting the outcome of any future vote.

.

McCain is a bad candidate with many negatives. If he keeps screwing things up at least some of the MSM will fall off his bandwagon. There will be a fall campaign and McCain is such a confused liar that a well run campaign should expose him.

SurveryUSA is supposed to be good. Gallop goes back and forth. I think Gallop sometimes polls to change people's minds, rather than measure what's going on. Zogby seems to be a complete sell-out since 2004. He used to be good. Got things the most accurate. No longer. SurveyUSA claims to be the most accurate, according to their own report cards... har har...

http://www.surveyusa.com/index.php/surveyusa-report-cards/

When I did the analysis of polls for the 2004 election, Zogby was by far the most pissy about being asked legitimate questions about how the poll was conducted. They seemed to be especially sensitive when asked to what extent the people who pay for the polls influence the questions asked and how the poll is conducted.

http://www.reuters.com/articlePrint?articleId=USN1824791220080319

Obama's lead over Clinton narrows: Reuters poll
Wed Mar 19, 2008 1:35pm EDT
By Steve Holland

WASHINGTON (Reuters) - Democrat Barack Obama's big national lead over Hillary Clinton has all but evaporated in the U.S. presidential race, and both Democrats trail Republican John McCain, according a Reuters/Zogby poll released on Wednesday.

The poll showed Obama had only a statistically insignificant lead of 47 percent to 44 percent over Clinton, down sharply from a 14 point edge he held over her in February when he was riding the tide of 10 straight victories.

Illinois Sen. Obama, who would be America's first black president, has been buffeted by attacks in recent weeks from New York Sen. Clinton over his fitness to serve as commander-in-chief and by a tempest over racially charged sermons given by his Chicago preacher.

The poll showed Arizona Sen. McCain, who has clinched the Republican presidential nomination, is benefiting from the lengthy campaign battle between Obama and Clinton, who are now battling to win Pennsylvania on April 22.

If we don't get our act together and pick a nominee, we're going to get creamed.

the dem chances to win back the White House. They have crusified Hillary on a daily basis, and now with this Obama crisis (which is huge and will not go away), it's time for the dems on this board to start facing some serious facts.

It's making me dizzy.

The song Dizzy in the 60's (or early 70's)

They used to play it at the Tilt-a-Whirl at the Calhoun County Fair, I think! (or it might have been the Scrambler). I can still summon up the memory of that greasy smell of oil from rides, and the music. I loved fairs!

Groups and individuals pay polling companies to conduct polls. Sometimes these contracting parties have an agenda.

For a poll to be valid it must be conducted in a way that will provide credible information about the group of people it is supposed to be studying.

American polling groups do not provide sufficient information to make this determination. I did an analysis of polling groups in the 2004 election and found widespread hinkiness. I contacted several groups and NONE would provide the requested information.

Issues that are important include:
1) What questions were asked?
2) Who determined the polling procedures (time of day called, questions asked, order asked, criteria for sampling).
3) Who is in control of what is released to the public? How it is released? When it is released? If it is released at all?

National Council on Public Polls has a good list of questions: "20 Questions A Journalist Should Ask About Poll Results" (http://www.ncpp.org/?q=node/4)

One thing you need to look closely for is if, when they give poll results, if they change the criteria of people questioned. For instance, there are two groups of people that are often polled: "registered voters" and "likely voters." If a question is asked of "likely voters," it increases up to 7% the conservative outcome of the answers (i.e., 7% more response for the more conservative of the options presented).

A CNN poll in 2004 pulled this trick to fraudulently present the results of a poll. The title of the article was "If Bush Ran Against Any Dem Candidate Today, He Would Win." They then asked a question for a choice between Bush Jr and each of the Dem candidates. For all of the Bush v. "Dem candidate" candidates, they asked of "registered voters." For Bush v.Dean (Dean was at that time, by far, the leader in the Dem race) they asked of "likely voters," giving an up to 7% more conservative outcome of the poll for that question. Of all the other Dem candidates v. Bush, Bush was far enough ahead that the 3.5% margin of error didn't mean anything. However, there was a 1-point lead between Bush v. Dean, which means it was well within the MOE AND, if you take in the switcheroo in the type of people they polled, it was likely that Dean actually would have won on that polling question.

Now, go back up and look at the title of the article.

Polls are tremendously powerful. Humans are herd animals and will move towards where the majority is moving. Political forces know this; hence, the polls.

Voters are being manipulated by the polls, and political forces are spending millions to do it.

And we can't even find out if the poll was done in an ethical matter or not.

Really fantastic.

You ought to make this its own thread, if you can yet.

Welcome to DU.

:toast:

I will dig up the analysis and write a more complete synopsis.

Perfect word for it.

Obama Clinton
March 19: 47 42
March 18: 45 44 -- Obama's speech
March 17: 46 44
March 16: 47 44
March 15: 46 45
March 14: 50 42
March 13: 48 41 -- Wright hits news
March 12: 47 42
March 11: 48 41 -- Mississippi
March 10: 46 44
March 9: 45 47
March 8: 45 46 --Wyoming
March 7: 43 49
March 6: 43 48
March 5: 43 48
March 4: 44 46 -- OH, TX, VT, RI

from Kos:

http://www.dailykos.com/storyonly/2008/3/19/133848/790/1009/480026

beat me to it
GRRRR :)

...at least on this subject.