| View previous topic :: View next topic |
| Author |
Message |
The Cyber Menace
Seaman Recruit
Joined: 22 Aug 2004
Posts: 49
|
 Posted: Thu Sep 09, 2004 6:41 pm Post subject: Dick Morris says Time and Newseek polls are accurate |
 |
|
http://www.nypost.com/postopinion/opedcolumnists/19281.htm
| Quote: | How big is Bush's lead? Don't believe the surveys that show it in the 5- to 7-point range. Believe the surveys of Time and Newsweek, which show a lead in excess of 10 points.
The difference is because pollsters disagree about whether or not to weight their results to keep constant the ratio of Republicans, Democrats and Independents in their sample. Some polling firms treat party affiliation as a demographic constant and, when they find that their sample has too many Republicans, they weight down each Republican interview and assign an extra weight to each Democratic response.
But other polling firms — and I — disagree. We feel that political party is not a demographic, like gender or race or age. If the survey finds more Republicans than usual, we think it's because the country has become more Republican, so we treat the result as a indicator of national mood, not of statistical error.
Time and Newsweek both picked up major moves toward the GOP in the wake of the convention. Likely the other firms did too, but they treated the finding as a mistake and weighed down the Republican interviews, making the race appear to be closer than it really is.
|
The only reservation I have about such analysis is the degree to which people really change party affiliation on the spot. Don't they need to be registered with a party to say they belong to that party? Even if they don't, I'm not sure if this trait is as variable as Morris and others believe.
Nevertheless, it can't be bad news for the Swiftees.  |
|
| Back to top |
|
 |
Arty Guy
Seaman
Joined: 20 Aug 2004
Posts: 190
|
| Posted: Thu Sep 09, 2004 6:45 pm Post subject: statistical interpretation |
|
|
| Actually the weighting described above makes a certain amount of sense to me. If a particular poll on a particular day happens to pick up a disproportionate number of people from a particular party, then the results are going to reflect this. Nevertheless, two of the polls show a markedly greater margin for Bush over Kerry. |
|
| Back to top |
|
|
The Cyber Menace
Seaman Recruit
Joined: 22 Aug 2004
Posts: 49
|
| Posted: Thu Sep 09, 2004 6:50 pm Post subject: Re: statistical interpretation |
|
|
| Arty Guy wrote: | | Actually the weighting described above makes a certain amount of sense to me. If a particular poll on a particular day happens to pick up a disproportionate number of people from a particular party, then the results are going to reflect this. |
I suppose that's where the dispute is. Are there more members of one party in a particular poll because of an error in sampling, or because more respondents now want to be members of that party? The first needs to be accounted for. The second is valid.
Nevertheless, four post-convention polls (all done by leftist media, mind you) show a healthy Bush bounce. |
|
| Back to top |
|
|
mj_brutus
Seaman Recruit
Joined: 24 Aug 2004
Posts: 32
|
| Posted: Thu Sep 09, 2004 7:02 pm Post subject: Weighting the results is just silly |
|
|
Imagine the question is: are you Rep or Dem?
The results come back 55 Republican 45 Democrat. But if one wieghts the results, it becomes 50-50. That is because, the assumption that the right "answer" is 50-50 forces the weighted result to *always be* 50-50, whatever the real world proportions are! |
|
| Back to top |
|
|
one more captins mast
LCDR
Joined: 10 Aug 2004
Posts: 438
Location: Texas
|
| Posted: Thu Sep 09, 2004 7:07 pm Post subject: thinking, thinking, thinking |
|
|
So the polls are what the lilttle man who wants to be vp, lawless
Edwards was crying about just before he came out and told
the REAL VP, "don't hit me any more, please no don't hit me
any more,"
If it was not so sad , it would be funny, as this means that
fully 30/40 % of our fellow Americans are "not in their
collective "right minds" for they "?support these two
strange men.
other than that have a nice day John / john
or in TV language. "good nite" John, "good nite" popa, , "good nite"
all.
_________________
the strange mr aj |
|
| Back to top |
|
|
Me#1You#10
Site Admin
Joined: 06 May 2004
Posts: 6503
|
| Posted: Thu Sep 09, 2004 7:19 pm Post subject: |
|
|
Admin note:
Please refrain from general political dialogue in the main forum.
This forum is reserved for discussion of John Kerry's unfitness for office.
We appreciate your cooperation in this regard.
This topic will be moved to Geedunk.
Thanks |
|
| Back to top |
|
|
Frank B
Ensign
Joined: 10 Aug 2004
Posts: 58
Location: Montana
|
| Posted: Thu Sep 09, 2004 7:42 pm Post subject: Da Polls |
|
|
"POLLS DON'T LIE.....PEOPLE DO."
IMHO, polls are not to be trusted. My questions about polls: How can a poll be "scientific??" How are people selected? How many are selected? Do their opinions represent the majority? What variables are plugged in? taken out? What guru(s) in what city behind what desk in what corner of the office determines what polls questions and how they are phrased?
Putting my sarcasm aside, common sense has to prevail when looking at polling results. For example: the U.S. has almost 300 million people. Poll numbers one week show Kerry ahead on any issue by 55 to 45. One week later after gas prices drop and the sun comes out, Bush is up 55 to 45. That 10% swing to Bush equates to 30 million Americans changing there minds!! Then the next week, Kerry picks up a couple poll points to make it 53-47....now 6 million of the 30 million changed their minds!! And those undecideds.....are we to believe these fence sitters have no views or opinions until right up to the last minute before the polls open?
This stuff will make you scratch your head bald.....and I don't have much to spare 
_________________
"Any government that believes money should be taken from Peter to pay Paul will, of course,
have the support of Paul" -- G B Shaw |
|
| Back to top |
|
|
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
