October 17, 2004

Hey you damn kids, get off my polling data! I think we all heard the thing about a month back, about how the election pollsters are missing the 5% of the population who have only cellphones and no land line (myself included) But what's shocking in this article is that the pollsters seem to blame the young people for not showing up in their data

"Many of these people are not voters," said Linda Piekarski, vice president of database and research at Survey Sampling International, which provides samples for the research industry. "They've always been hard to get into our polls anyway. They tend to be non-responsive." Nice broad generalization, backed up by absolutely no data, from someone in an industry based on stats! Excuse me but as a pollster, isn't it your job to get everyone into your data, whether it's "hard" or not??

  • just curious, as a follow-up, how many of us here are "cell phone onlys Has anyone hear ever actually talked to a pollster? I haven't, even when i had a landline.
  • I'm in the cell phone only camp, landline-free since 2001. The last time I talked to a pollster was in 1987, when I had my first job in high school working nights as a pollster. I just realized that I only have the cell phone numbers for most of my friends and family. Most of the people I know that still have regular phones have them hooked up primarily for faxing or backups when the cable 'net connection goes kerplooey.
  • I've been cell phone only since 1999 and I've never spoken to a pollster and only twice to a telemarketer. Life, without a landline, is truly beautiful. Oh and I vote. Lets hope all of those rebelious enough to give up the tether are right there with me.
  • I do not have only a cell phone, as I hate them. And I have never spoken to a pollster. I can say that if I ever do, I plan to give completely bogus answers to everything. I think that the way polls dictate the way other people vote is absolutely frightening. If it became widely known that a large group of people were giving dishonest answers to these questions, then people would be less likely to vote for frontrunners on election day. Do not underestimate the amount of people who only want to vote if they think that their candidate is winning. My solution? Render the polls meaningless.
  • When I had a landline, I was something of a poll-taking whore. I must've gotten on a list or something, cause I kept getting calls from both marketers and politikers. Now that I'm wire-free... bupkis. Polls are pretty worthless under a microscope. Poking holes in the methodology is pretty easy. I've got my saltshaker at the ready everytime I read one.
  • speaking of REALLY young people, the weekly reader poll has correctly predicted the american election beforehand (if i remember) every year going back several decades. it'll be announced oct. 25.
  • dogburp is right about the hole in the methodology, but consider this: exit polls conducted by the major US news organizations at election cites are VERY accurate. They had spent a great deal of time and money over many years to get good results. They were correct about Florida in 2000. But the stealing of the 2000 election has managers and executives of these news organizations doubting the validity of their own polls now. It is more likely this time than last that an election "result" differing from the polls will be accepted. This makes it all the easier for anyone stealing the election to get away with it.
  • I have a landline but don't answer calls unless I recognize who's calling. Thus we're certainly not getting polled either.
  • Before I got my landline here, the Electoral Register actually sent a nice lady here a few times in the evenings to tell us Commonwealth citizens that we were eligible to vote! w00t! She explained that they had to contact us so that we could tell her that we didn't want to be on the "junk mail" list, which allows anyone who asks full contact information about us. Before that, in Singapore, nobody bothered with polling. Half the constituencies are walkovers for the dominant party, and the other half are about 80-90% in the bag for the dominant party. Although my constituency was the one with the highest number/percentage of spoiled votes *grin*. So... when's the election in UK taking place? /ignoramus
  • Never had a cell phone, and I don't want one. I don't answer the phone (usually) unless I recognize the caller's number or name. Some cell phone numbers show up as unknown on the caller ID, so sometimes I answer the unknown calls, thinking it might be one of my kids. So this year alone, I have been pollled three times. All three times the number was unknown by my caller ID.
  • Before that, in Singapore, nobody bothered with polling Erm, er, no.
  • I think this article makes a very valid point and I think that there will be much post-election analysis of how the polls could have been so wrong. Assuming that the voters votes are actually counted (a big assumption after what we saw in 2000 in Florida) I don't see how this election is even going to be close: I think that the huge increase in new-voter registration points to large numbers of people being activated by discontent with the present. (It is an assumption that the majority of these are voting against Bush, but a pretty safe one - i think. I also think that people forget how many people voted for Gore (or against Bush) in the 2000 election. I have a hard time believing, particularly in the face of evidence such as those of lifelong Republican's voting against Bush this time around, like the son of former Republican U.S. President John Eisenhower that the Republicans have lost a significant percentage of their base. It is remarkable how anti-conservative in agenda the current bunch is. I also think it will be extremely funny the reaction of the media will be once the fear of retribution from the current administration is lifted. I think you will be amazed by the turns taken in the coverage given to the Plame investigation and a multitude of other stories that are being criminally under-reported in the current environment.
  • It would be nice if it were so, mecurious, but that optimism is unwarranted, in my opinion. The media is shit; that will not change until the FCC grows a pair or people start leaving TV in droves. The former doesn't seem likely to me, though the latter by all accounts is well under way. My prediction is a decisive Kerry win, but a hung senate and a Republican house. A Republican senate with Specter as chairman of the Judiciary Committee wouldn't be the end of the world either. (Specter won't let Roe/Wade overturners through. I hope.) Until 2006 there will be no real house-cleaning. That election will determine if we can start seeing any actual progressive stuff. I would love to be wrong as long as things turn out more according to your predictions. This is the minimum I can tolerate.
  • In the polls The New York Times calls the national horse race a tie at 46 percent among registered voters. That's good news for Kerry, as the challenger traditionally picks up a majority of the undecided voters on Election Day. Zogby's tracking poll concurs with the Times, finding both candidates tied at 45 percent for the second straight day among likely voters. Rasmussen's tracking poll isn't too far off from that, giving Bush a 48-47 lead. And exactly two weeks before the election, the Times reports that Bush's disapproval ratings are near an all-time high. Only 43 percent of those surveyed approved of the president, 45 percent disapproved. That number isn't much worse than Kerry's, who earned disapproval from 44 percent of voters. But while Kerry's own numbers certainly aren't reason for him to celebrate, he should be thrilled with Bush's: In recent elections, an incumbent candidate's share of the vote almost never exceeds his approval rating on Election Day. Bush also stacks up poorly against another candidate -- himself. In October of 2000, voters surveyed by the New York Times approved of the then Texas governor by a 48-33 margin. In state-by-state polling, Survey USA reports that Kerry now leads in two out of three of the big swing states: He's ahead 50-49 in Florida, and by a 51-45 margin in Pennsylvania. And though recent New Jersey state polls have shown that Bush is surprisingly competitive, a Rutgers-Eagleton Ledger poll thinks he's wasting his time: Kerry leads 51-38. Finally, a Newsweek poll shows young voters favor Kerry, though not overwhelmingly. Kerry leads among all voters aged 18-29 by a margin of 47-45, though his advantage increases among likely voters to 52-42. One reason that the youth turnout might be high this year is that more than a third of them expect a draft to come back. 34 percent said it was at least "somewhat likely" that the United States would be returning to conscription as a result of the Iraq war. from the 10/19 Salon war room page I stand by my prediction.