posted
Has anyone else noticed the strange behavior of the Zogby polls? Due to the methodology of the site and the increased activity of the pollster, Zogby polls dominate the map currently.
Early on I suspected Zogby of partisanship and manipulation because they seemed to consistently skew towards Kerry. Then I looked at the 2000 numbers and found the same trend, with their polls favoring Gore more than anyone else. Their polls, however, more closely reflected the outcome than any other pollster. I cut them some slack, but now I am not so sure.
They may have changed their methodolgy or something as their results now seem more erratic than anything. They still skew to the left in some states, most notably Colorado. However, their New Mexico, Ohio and Florida polls skew very far to the right. In Florida, the difference between Zogby and the mean of the other polls is outside the margin of error. I don't really know what's going on here, but it does make electoral-vote more interesting.
Posts: 1769 | Registered: Feb 2000
| IP: Logged |
Most polls are done over the telephone, limited to landlines.
More people this year than ever before have gotten away from landlines, and are using mobile phones as their only phone.
The result is that this segment of society is being entirely missed by the pollsters.
Democrats are hoping that those being missed are mostly poor or young or Tech-obsessed who all leaning to democratic.
Republican are hoping that young, poor, and tech-obsessive are not poeple who normally vote anyway, so it won't make much of a difference.
The Pollsters are quietly hoping that the randomness of the situation won't effect their margin of error.
Posts: 11895 | Registered: Apr 2002
| IP: Logged |
posted
It's not just that, Dan. I think both the Repubs and the Dems are very interested in seeing incorrect results in polls, both pertaining to themselves and for the other party.
Posts: 13123 | Registered: Feb 2002
| IP: Logged |
posted
Studies have been conducted of those who use only mobile phones. They found these people were generally young, many were students, and among those registered to vote, they 50% favored Kerry while only 30% favored Bush.
That said, I'm not seeing the polls as terribly reliable right now. When they disagree with each other way outside the stated margin of error, something is definately wrong.
Posts: 1769 | Registered: Feb 2000
| IP: Logged |
posted
As a resident of Wisconsin, I've had four phone polls, two postal polls, three E-mails, and three door-to-door polls. So far. *shudder*
Posts: 37449 | Registered: May 1999
| IP: Logged |
posted
"Skew"... from what? Are you privy to the actual results?
We are mainly on cell phone, and not likely to vote Kerry. Though in general, I've expressed before that young people are more inclined to be democrats.
Posts: 11017 | Registered: Apr 2003
| IP: Logged |
posted
Yes, the cell phone/land line bias is very interesting. I wonder if it is enough of a difference to spoil samples?
Posts: 6316 | Registered: Jun 2003
| IP: Logged |
posted
Pooka, I meant relative to the mean numbers from all the pollsters.
If you look at the per polster data on Electoral-Vote, it will be pretty clear what I mean. Zogby isn't the only one that has a noticable bent towards one candidate. Strategic Vision is openly partisan, and its polls reflect that. They are probably the most heavily and consistently skewed, always to the right, and very frequently they disagree with the mean by greater than the stated margin of error. Gallup also tends to lean right, but mostly within the margin of error. This could be the result of different methodology or just random chance.
Posts: 1769 | Registered: Feb 2000
| IP: Logged |
posted
Not having a phone, I don't get those calls. (Also in Wisconsin). My mother, on the other hand, has a riot with the poor survey people. This is truth: She supports Bush, Feingold, and owns a handgun. There are few people who can figure that out, and none of them work for the polling companies. She spent half an hour on the phone with one guy listening to him saying "I don't get it!" and not explaining anything. Another time last week, she spent at least fifteen minutes with one of those polls where they ask how strongly you feel about an issue. When she wasn't answering "Oh, I feel very strongly," she was asking what her options were.
posted
What good is a strictly partisan poll. I guess if you're trying to decide how many voters are going across party lines then maybe it would be interesting. But in general what good is a poll that isn't done accurately?
Posts: 872 | Registered: Mar 2002
| IP: Logged |
posted
My Government teacher said that Zogby uses cell phone numbers as well as land line numbers, and that is why his results are different.
Posts: 1015 | Registered: Aug 2004
| IP: Logged |
posted
I get polled all the time but mostly for useless things like local measures. They prolly keep calling me becuase they know I'll give them my opinion. (though you'd think they'd get bored of my lectures by now.)
Posts: 7085 | Registered: Apr 2001
| IP: Logged |
posted
I want to write the polls. I think trying to make completely unbiased poll questions would be really fun.
Posts: 4655 | Registered: Jan 2002
| IP: Logged |
posted
Adrianna Whats-her-name is starting a campaign to get people to stop talking to pollsters. I just caught the tail end of a bit on NPR, she was saying if enough people stop talking to them to get the response rates in the single digets everyone will have to acknowledge that the polls are worthless. I don't remember why in particular she thought this would be a good thing, but I'm sure someone who's more curious than I am right now will find the website and let us all know.
Posts: 7954 | Registered: Mar 2004
| IP: Logged |
posted
I've never been polled by a legitimate political polling company. I was polled by some left wing group with very biased questions once...but never by a non partisan group. I have also gotten polls from companies and such about different products...some of those are balanced...and sometime you can tell that they are trying to skew answers a certain way