Pollsters' Inaccuracy (Zogby) In 2002 Contests Worries Industry

Drudge Report ^ | Nov. 8, 2002 | Matt Drudge

[Pres Raygun]

Most people do not understand the difference between precision and accuaracy. Precision refers to how closely repeated polls of the same question and sampling technique will get the same answer again and again. Precision says nothing about whether the poll is correct. Accuarcy is how closely the poll come to reality. When pollsters give their poll's margin of error they are talking only about the poll's precision, while most people are thinking they are talking about accuracy. A poll can have a margin of error of plus or minus 5%, but be inaccurate by over 10%, because the pollster did not poll an unbiased sample. To fix this problem the pollsters fudge their poll numbers by looking at the results of previous elections compared to their poll numbers for those elections. That works unless something happens since the last election that changes the political landscape, such as 9/11, W camapigning, RATS in Bagdad, etc. Polls are still very good at showing change within a campaign. Even though their accuarcy may be off significantly, a significant movement in a particular poll number (taking into account the precision of the poll) largely is reflecting a real change in the voters intentions.

[PolishProud]

According to a Fox News election day poll regarding the Allard-Strickland race, newly arrived Liberal voters canceled out native or long time Colorado conservative voters, but senior citizens went 53 to 46 for Allard.

Does anyone know how senior citizens voted nation wide?

[SpinyNorman]

It was the lack of media manipulation, usually backed by the VNS. The media has been whining about how this election was reported with 1969 standards and how their viewers deserve or demand the fastest possible results, blah, blah, blah.

I thought that the news was, for once, refreshingly devoid of the usual projections designed to manipulate voters and opinions. Why can't we wait for the actual votes to be counted. Given the closeness of races like the Johnson/Thune race, projections and predictions can be detrimental in that an opponent may concede too early, a la Gore (okay, not really!!), or worse, give the Dems the info they need to manufacture sufficient votes to squeak by like they did in South Dakota. I am sure that we all noticed how the corrupted Indian precincts were the last to straggle in, even though that has not been the norm in previous elections (Daschole's last election to name one). Clearly, they were digging up (pun intended) the necessary votes to swing the election.

[TroutStalker]

Here is the article from the WSJ that this story is based on:

Why Some Pollsters Got It So Wrong  ^

[annieokie]

Year ago, when LBJ was in office, my dad got a TM call asking for a donation. My dad told them to call the President as he does all his donating. LOL

I tried this same thing when the Toon was in office, I simply said to call 1600 Penn Ave, as they did all my donating for me, the TM caller asked me for the phone #, ROFL. When she finally got the picuture after several other stupid questions about 1600 Penn ave., I then had to tell her IT IS THE PREISDENT OF THE UNITED STATES, click-hangup.

[Salvation]

Zogby's polls always lean to the left because that is whom he chooses to poll. Simple........

[BlackbirdSST]

I've never had a very high opinion of poll's, pollster's, or people that stick their finger to the wind before they form an opinion. I'm sure I'll get plenty of response from freeper's about how accurate statistical data is and blah blah blah, but this is ample proof that polling is nothing better than a W.A.G.! I can offer just as much to the world of finger in the wind types, what's it pay? Blackbird.

[pabianice]

And so, it appears, did many other political pollsters. The reasons may be as various as the recent popularity of caller ID and cellphones, which hamper efforts to reach voters, and the nation's increasing ethnic diversity, which makes it harder to get an accurate statistical sampling of the electorate."

Bingo! A winner! Basically, most people who answer their home phones during the day are people on welfare...

[pabianice]

"People lie." I was polled on the phone on 11-4-2002 I lied my ass off! As far as the pollster is concerned I was for divided government, not invading Iraq, and a Democrat majority in the Senate. Bwah ha ha ha!

BTW, I'm the "102-year-old female from Michigan" who logs into Washingtonpost.com...

[FreedomPoster]

Yes, but if they can't reach them because their "out of area" call is call-screened. . . .

[Republican Red]

I remember Roger Ailes(of Fox News fame)explained the failed polling process the best. He stating that when a democrat is approached by a pollster he will pour his heart out about his "feelings" on the matter. The Republican will tell you "it's none of your business" and walk off.

[Rodney King]

The country's continuing stream of immigration also makes accurate polling more difficult, since racial and ethnic groups tend to have distinctive voting patterns.

That they have distinctive voting patterns ought to make prediction easier. The real issue here is caller ID. Conservatives use it more than liberals do.

[DUMBGRUNT]

The FTC charged Miss Cleo with deceptive advertising!
Pollsters/Miss Cleo same/same!