I’m not certain I understand your argument … are you saying that those contacted by gallup who identify themselves as republican or democrat should automatically vote for their party’s candidate? And gallup isn’t purposefully trying to contact an even number of republican and democrat candidates such that the voting sample is 50/50, they contact random samples such that the sample ends up being 7% or 12% or whatever % republican + or – the margin of error. If anything about the gallup polls or any other telephone polls are off it’s the fact that they do not contact cell phone users …

___________________
Retro ...

i think his arguement is that a credible poller should not use a 40% or higher republican base. the exit polls in 2000 had a 39% democrat and 35% republican turnout. it has nothing to with if whether people vote for their party... that's what the poll is for. In order to paint a clear picture on how an election will turn out, they should at least get their initial figures correct.

Good post, Opus

edit: i just visited electoral-vote.com and he has the same story posted. he also provided a link for anyone who wanted to argue exit poll figures:
http://www.udel.edu/poscir/road/cou...pollsindex.html

he went as far as to say that most pollers are swinging toward the direction in which the people who pay their bills want them to swing. i guess we'll have to wait until november to see how true this really is.

Last edited by ResonantDrag on Sep-28-2004 at 17:43

quote:

Originally posted by ResonantDrag i think his arguement is that a credible poller should not use a 40% or higher republican base. the exit polls in 2000 had a 39% democrat and 35% republican turnout. it has nothing to with if whether people vote for their party... that's what the poll is for. In order to paint a clear picture on how an election will turn out, they should at least get their initial figures correct. Good post, Opus

Using bases based upon previous elections don't quite paint an accurate picture of what the turnout will be in a heated election such as this one. Gallup describes its sampling methadology:

quote:

Results based on likely voters are based on the subsample of 758 survey respondents deemed most likely to vote in the November 2004 general election, according to a series of questions measuring current voting intentions and past voting behavior. For results based on the total sample of likely voters, one can say with 95% confidence that the margin of sampling error is ±4 percentage points. The likely voter model assumes a turnout of 55% of national adults. The likely voter sample is weighted down to match this assumption.

So ultimately however people answer the questionaire determines their "liklihood" to vote. Whether people actually do as they say they do on the questionaire is anybodies guess but there's not really any other better method to use and that's hardly gallup's fault.

___________________
Retro ...

Last edited by occrider on Sep-28-2004 at 17:51

quote:

Originally posted by occrider Using bases based upon previous elections don't quite paint an accurate picture of what the turnout will be in a heated election such as this one. Gallup describes its sampling methadology:

True, but their methodology should come into question with such a discrepency in likely voters this year compared to the trend in the last 3 elections. Coupled with the fact that new voter registration for Democrats is greatly outpacing Republican registrations, I'd venture to say their methodology merits some scrutiny.






Edit: perhaps I'm being more speculative than concrete with my contentions here, and I'll admit as much. But can anyone truly believe that the GOP has a 12% higher likely voter turnout over the Democrats? Just that alone seems pretty questionable.

___________________
Whence September dusk grows crisper still,
with leaves all crimson conquered,
I yearn to shout,
and dance about,
and stick pickles in my honker...

quote:

Originally posted by MisterOpus1 True, but their methodology should come into question with such a discrepency in likely voters this year compared to the trend in the last 3 elections. Coupled with the fact that new voter registration for Democrats is greatly outpacing Republican registrations, I'd venture to say their methodology merits some scrutiny.

Yes but looking back at previous turnouts is hardly "predictive" in nature . Like I said if these people fill out on the questionaires that they are definetely going to vote or whatever the question is, it's somewhat difficult to discount that voter because your overall GOP voter turnout number isn't kosher with what it was in the last election. As for new voter registration, keep in mind that Gallup polls only contact participants with telephones. Perhaps many of these new registrants are 18-24 year olds or whoever and they do not have or really use their telephones?

quote:

Edit: perhaps I'm being more speculative than concrete with my contentions here, and I'll admit as much. But can anyone truly believe that the GOP has a 12% higher likely voter turnout over the Democrats? Just that alone seems pretty questionable.

No not really ... but you never know, maybe they rallied all the religious kooks. What does zogby or harris say?

___________________
Retro ...

quote:

Results based on likely voters are based on the subsample of 758 survey respondents deemed most likely to vote in the November 2004 general election, according to a series of questions measuring current voting intentions and past voting behavior. For results based on the total sample of likely voters, one can say with 95% confidence that the margin of sampling error is �4 percentage points. The likely voter model assumes a turnout of 55% of national adults. The likely voter sample is weighted down to match this assumption

i just don't see how this justifies such a large shift in expected party representation. there's been a lot of studies made with minority groups such as blacks, asians and latinos and they're expected to turn out in record numbers (i doubt bush's plan B will work as well this year). As this election heats up, voters will still continue to register. I think the 29th is going to be a large student push for registration, and that's rarely good news for republicans. Oct. 12th is the last day to register in my state, but i don't know if that's a national figure.

to give 95% certainty of turnout on polling 758 people is rather absurd, especially in this election. you did make a very good point on cell phone users, as i'm sure that they vote as well. I think that with the scant evidence on any real party shift, they should be playing it safe and keeping the ratio as close to 2000 as possible. unless there's some other source than a questionable organization's questionable justification for it's "normalization" of data.

here's something to chew on... i'm still sorting out some usable details.

http://www.eriposte.com/election04/...ing_polling.htm

quote:

Originally posted by occrider Yes but looking back at previous turnouts is hardly "predictive" in nature . Like I said if these people fill out on the questionaires that they are definetely going to vote or whatever the question is, it's somewhat difficult to discount that voter because your overall GOP voter turnout number isn't kosher with what it was in the last election. As for new voter registration, keep in mind that Gallup polls only contact participants with telephones. Perhaps many of these new registrants are 18-24 year olds or whoever and they do not have or really use their telephones?

I think the main gripe lies in the telephone answers, at least this is what Zogby has stated in the past. I agree that it's not a very strong trend to examine the last 3 elections and base a conclusion on the current Gallop polls on that - which was why I qualified my statements with the edit in the last post. However, I do believe that they should be reexamining their methodology. Didn't Gallop have Gore down by like 7% points the day before the 2000 election, only to have Gore win the popular vote by 500k?

quote:

No not really ... but you never know, maybe they rallied all the religious kooks. What does zogby or harris say?

Argh, didn't have time to examine them fully, but they appear at quick glance to have it much closer between the two parties:

http://www.harrisinteractive.com/ha...dex.asp?PID=495

http://www.zogby.com/news/ReadNews.dbm?ID=867

Must eat now. Energy low. Crankiness high...

___________________
Whence September dusk grows crisper still,
with leaves all crimson conquered,
I yearn to shout,
and dance about,
and stick pickles in my honker...

quote:

Originally posted by ResonantDrag i just don't see how this justifies such a large shift in expected party representation. there's been a lot of studies made with minority groups such as blacks, asians and latinos and they're expected to turn out in record numbers (i doubt bush's plan B will work as well this year). As this election heats up, voters will still continue to register. I think the 29th is going to be a large student push for registration, and that's rarely good news for republicans. Oct. 12th is the last day to register in my state, but i don't know if that's a national figure.

What is the propensity for students or minority groups to respond to pollsters from their home telephone? Additionally perhaps students and minorities are an outlier for this election. I'm not saying that the gallup poll is an accurate representation, I'm merely justifying the findings. I have a funny feeling that this entire election is something of an outlier so I'm not entirely sure what to expect.

quote:

to give 95% certainty of turnout on polling 758 people is rather absurd, especially in this election. you did make a very good point on cell phone users, as i'm sure that they vote as well. I think that with the scant evidence on any real party shift, they should be playing it safe and keeping the ratio as close to 2000 as possible. unless there's some other source than a questionable organization's questionable justification for it's "normalization" of data.

Hey, don't knock the central limit theorem . It kicks all ass. If that 758 people is a random sample representative of the voters in this country than that's how the country is going to vote. I would argue that the 758 people may not be accurate due to some factors I outlined above. But also bear in mind that the margin of error is +-4 points meaning that Bush's lead is not statistically significant.

___________________
Retro ...

quote:

In 3 of the last 4 presidential elections (including the last one), Gallup's final RV (registered voter) reading was actually closer to the final result than their final LV (likely voter) reading!

quote:

Here is a summary of the case against Gallup's LV data: ...Gallup decides who likely voters are based on 7 questions about their interest in voting, attention to the campaign and knowledge about how to vote (e.g., where their polling place is located). The interested/attentive/knowledgeable voters are designated “likely� and the rest are thrown out of the sample. But as a campaign progresses, the level of interest among voters tends to change, particularly among those with partisan inclinations whose interest level will rise when their party seems to be mobilized and doing well and fall when it is not. Because of this, partisans of the mobilized party (lately, Republicans) tend to be screened into the likely voter sample and partisans of the demobilized party (lately, Democrats) tend to get screened out. But tomorrow, of course, the Democrats could surge, in which case their partisans may be the ones over-represented in likely voter samples. That suggests the uncomfortable possibility that observed changes in the sentiments of “likely voters� represent not actual changes in voter sentiment, but rather changes in the composition of likely voter samples as political enthusiasm waxes and wanes among the different parties’ supporters. And that is exactly what political scientists Robert Erikson, Costas Panagopoulos, and Christopher Wlezien find in their analysis of Gallup's 2000 RV/LV data in their forthcoming paper, “Likely (and Unlikely) Voters and the Assessment of Campaign Dynamics� in Public Opinion Quarterly: “shifts in voter classification as likely or unlikely account for more observed change in the preferences of likely voters than do actual changes in voters’ candidate preferences.� That means that, instead of giving you a better picture of voter sentiment and how it is changing than conventional registered voter data, likely voter data give you a worse one since true changes in voter sentiment are swamped by changes in who is classified as a likely voter. ...And then there's this: the LV data haven't been working so well lately even right before the actual election. In 3 of the last 4 presidential elections (including the last one), Gallup's final RV reading was actually closer to the final result than their final LV reading!

quote:

1. The latest Gallup Poll has Bush ahead of Kerry by 52-45 percent among likely voters but by only 49-48 percent among registered voters. Based on the numbers of registered and likely voters in their sample, this means that Gallup is projecting that 89 percent of Bush supporters will vote but only 79 percent of Kerry supporters will vote. That seems unrealistic. It is way out of line with data from the American National Election Studies on turnout among registered Dems and Republicans in recent elections. For the past three presidential elections, the turnout gap between Republicans and Democrats has averaged 3 percentage points and was never larger than 4 percentage points. The smallest gap was in 1992 (1 point), the election with the highest overall turnout. Assuming that 2004 will be another relatively high turnout election, we should probably expect a relatively small turnout gap, similar to 1992.

quote:

But if the party identification distribution is fairly stable and tends to change rather slowly, why would polls suddenly be turning up unrealistically high numbers of Republican identifiers? The best explanation, in my view, is that when the political situation jazzes up supporters of one party, they are more likely to want to participate in a public opinion telephone poll and express their views. An increased rate of interview acceptance by that party's supporters would then skew the sample toward that party without the underlying distribution having changed very much, if at all.

the rest of this article continues to reinforce the position that there is something fishy with the normalization techniques used nowadays. i'd prefer not to cut and paste the entire thing.

quote:

Originally posted by occrider Hey, don't knock the central limit theorem . It kicks all ass. If that 758 people is a random sample representative of the voters in this country than that's how the country is going to vote. I would argue that the 758 people may not be accurate due to some factors I outlined above. But also bear in mind that the margin of error is +-4 points meaning that Bush's lead is not statistically significant.

i am but a lowly mortal, i would never knock the central limit theorem. i just feel that to take in account the outlined factors, the error would be much greater than +-4 points. that has nothing to do with the law of averages, rather the pollsters ability to understand numerable factors.