Rasmussen SUNDAY Poll

Sunday September 05, 2004--The Rasmussen Reports Presidential Tracking Poll shows President George W. Bush with 48% of the vote and Senator John Kerry with 46%. The Tracking Poll is updated daily by noon Eastern. Two-thirds of the interview for today's report were completed after the President's speech on Thursday night.

Over the past nine days, Bush has been ahead eight times and the candidates have been tied once. Senator Kerry has not been ahead in the Tracking Poll since August 23.

Something's gotta be wrong here. There are two polls that contradict this beyond the MOEs.

The Newsweek poll is closer to RASS if you look at the proper weightings - The only reason the Newsweek poll shows such a large lead is do to the over-sampling of REP's -

Now I do think Rass is skewed a little toward the Dem's as well - I think GWB has a 5 to 8 pt lead right now -

If Rasmussen is getting a 39-34-27 split on a truly random basis then fine. If he is continues to poll until he gets the 39-34-27 mix then obviously the polling is skewed, biased and not valid.

This is exactly what I was trying to say. Once pollsters start "weighting" the sample in accordance with preconceived assumptions about party affiliation, the poll is no longer random and, therefore, no longer scientific. The best polls are the ones that correctly identify and randomly sample likely voters. Period. If this means more Republicans are included because they are more likely to vote than Democrats, too bad. Or vice versa. That's why I trust Gallup and almost no one else for top line numbers.

I'm afraid Rasmussen is setting his party affiliation dynamics to reflect 2000. He could be setting himself up for even worse results this time by weighting. We don't know what the percents will be that is why we have elections. Rasmussen is saying Democrats will win every time with his numbers and no Republican will ever be able to win in a landslide. Post 9/11 I think he may just be off in his analysis.

I agree. He got badly burned in 2000, and now he may be overcompensating in the other direction. And yes, 9/11 changed everything. National security is the 800 pound gorilla of the 2004 election, and on this issue every poll -- the good, the bad, and even the ugly -- shows Bush with a huge advantage over Kerry on this issue.

Food for though. If the media is biased, and the media has an agenda, and the media uses their front page reporting to forward their agenda, wouldn't the polling conducted by those organizations be suspect as well? If I were the head of CNN, Time, Newsweek, or the like, wouldn't I use my polls in the same way that I use my reporting? I can't wrap my mind around some vast conspiracy theories where showing a big bounce for Bush really helps Kerry in the long run. But seriously, if I don't trust these guys to give me the straight story on national events, why would I trust their polls?

Agree with your premise here - No doubt the networks USE polls when they can - and they have their polling companies push results they want -

With that said, I think if you break down the NewsWeek and the CNN/Time polls....GWB has a clear lead at this moment...but it is more around 5pts -

Perhaps this is the LIB media trying to force Kerry to get more aggressive (even though he has been) - Perhaps the media is trying to show a larger bounce....so they can make more news about "Kerry storming back" -

Who knows -

But what I do know is we need to stay consistent - We have to look at polls factually and when polls are weighted by over-sampling REP's .....we have to say these polls are not accurate -

The Newsweek poll is closer to RASS if you look at the proper weightings - The only reason the Newsweek poll shows such a large lead is do to the over-sampling of REP's -

I don't think we have enough information yet to conclude that for two reasons:
The Time poll mirrors the newsweek poll; so you have to establish that something is wrong with BOTH of those polls.
Secondly, whether or not they polled more Reps than Dems is irrelevant unless we know the proportion of Reps and Dems who will be voting on Nov 2; i.e. they need not be equal.

I don't think we have enough information yet to conclude that for two reasons: The Time poll mirrors the newsweek poll; so you have to establish that something is wrong with BOTH of those polls. Secondly, whether or not they polled more Reps than Dems is irrelevant unless we know the proportion of Reps and Dems who will be voting on Nov 2; i.e. they need not be equal.

Well, I think the point here is that we do know that both the NewsWeek poll and the Times/CNN poll oversampled Rep's compared to their previous polls.....and also compared to typical voter breakdown on past elections days -

We also know though that the MO is on our side - and that it appears if you use a "normal" voter break-down that GWB is leading by 5pts or perhaps a little more -

We have to be willing to admit that - and still break polls down that show GWB in the lead with proper weightings -

The Rass poll seems a little heavy for Dem's....but the Newsweek poll is definitely heavy for Rep's -

The Newsweek breakdown was 37D-31R (or close to that a fellow freeper dealing with polls showed) - This is WAY oversampled in favor of Rep's - (we can't deny this) - a consistent breakdown on the national level is right around 39D-35R (or maybe 36R) -

Factually incorrect. The sampling for party affiliation should reflect the actual distribution in the voting population. We know that there are more Reps than Dems in the U.S. today. See "Vital Statistics on American Politics". Current statistics also indicate that there are more people now who regard themselves as conservative than liberal (same source). This trend toward conservatism started after Reagan's election in 1980. So, no, I don't think we can assume what you are assuming.

"Food for though. If the media is biased, and the media has an agenda, and the media uses their front page reporting to forward their agenda, wouldn't the polling conducted by those organizations be suspect as well?

If I were the head of CNN, Time, Newsweek, or the like, wouldn't I use my polls in the same way that I use my reporting?

I can't wrap my mind around some vast conspiracy theories where showing a big bounce for Bush really helps Kerry in the long run. But seriously, if I don't trust these guys to give me the straight story on national events, why would I trust their polls?"

Sure they do. And I can show you how CNN, Time etc will use this BIG BUSH lead.

They will run with the double digit Bush lead. Then they will show much closer to the election a slight Bush lead but just a few points....clearly showing Kerry is gaining strength. Then right before the election they will report Kerry with a 7 point lead...clearly showing he's the winner and all the bandwagon (you know Iowa, NH and SC type voters...and there are MANY) voters will think Kerry is the winner and you just swayed a nation to vote for John Kerry...not through anything he did but rather the perceived image the nation thinks he is the winner.

The unfortunate thing is news companies commissioning a poll to MAKE news. This is very disturbing actually. They should be reporting on news that happens...not creating news to report.

Remember whenever you see a MSNBC, CNN, USATODAY, ABC, NBC, CBS, FOXNEWS, etc....in a poll it has a bias built into it by the news agency itself. That is one reason I have always like Rasmussen but I think his methodology is a bit off.

Factually incorrect. The sampling for party affiliation should reflect the actual distribution in the voting population. We know that there are more Reps than Dems in the U.S. today. See "Vital Statistics on American Politics". Current statistics also indicate that there are more people now who regard themselves as conservative than liberal (same source). This trend toward conservatism started after Reagan's election in 1980. So, no, I don't think we can assume what you are assuming.

There has never been a national election in American History since 1960 where on election day....the actual voter turn-out has showed MORE Republicans then Democrats -

Those are the real facts...and not some educational analysis of the subject - Just the cold hard facts - For example in the 2000 election - The breakdown was 4 to 6 % more Dem's then Rep's voted!

As for the trend of Conservative Vs Liberal - I'd agree more people would call themselves conservative ...but that means nothing ....other then "liberal" has a negative conotation to it -

The afrian-american population routinely call themselves conservative in polling after polling....Yet they vote 91% for the Democrat candidate -

No. You're quite wrong. Rasmussen is not a Republican. He is non-partisan. I believe that after publishing a few Republican-leaning polls, and getting scads of attention from folks like Limbaugh, he began publishing what his audience, Republicans and Freepers, wanted to hear. He was a little nobody who found himself getting millions of snoopers, until he was shown to be dead wrong. Then he lost his ties to conservatives. His track record is lousy, and no-one would pay any attention for him, other then his automation gets results first.

Zogby is not a Democrat, either. He got big by doing polls for the New York Post, once it was taken over by the same company that runs Fox News. In 1993 and 1994, he was very influential in bringing Republicans to power in New Jersey and New York, such as Whitman, Pataki, and Giuliani, by publishing polls which showed those races as rapidly tightening. In fact, Murdoch was pushing Zogby's polls so hard, he cut the price of the POst in half, and then offered it free.

Then Zogby went national with Fox News, who dropped him.

Zogby always used innovative formulae for predicting races. And that's what his pools always were: poll-based predictions, not polls.

Stick to the pros: Gallup's surveys of REGISTERED voters. (Even Gallup has started corrupting their polls with prediction methods, hence, Gallup's "likely voters" polls. These polls are ridiculous. If you announce that Bush is up by 6 among likely voters, and down three among RVs, and liekly voters are 90% of your survey, what percentage does he get among the registered-but-not-likely voters? 10%? Fortunately, Gallup has money to spend, and does its best work in pre-sampling to make sure that its raw (RV) polls are excellent.)

Just the oppposite. Likely Voters polls are cooked polls. Every pollster relies on correct predictions for attention, so he adjust his polls to fit his own expectations, and calls this "likely voters." But the real data is Registered voters.

Problem: many polling outfits simply pull "registered voters" at random, and then take guesses at why their data doesn't fit expectations. So cheap outifts, like Rasmussen, etc., know their RV data is so bad, they don't bother publishing it, then they make guesses. It's really no more useful that the Iowa markets.

If you want to REALLY know polling data, look at RVs of a long-established, commercial polling firm, like Gallup. Forget academic firms; their sampling is usually atrocious. Forget newbies, like Rasmussen and Zogby; they use "innocative" unproven methods at guessing. Forget polls too strongly associated with a single client, like Yankelovich (Time); they are pressured to guess by clients who don't like surprises and will, out of ignorance, blame them for statistical outliers.

You may be right on that point, but it still doesn't matter. You nor I know why Newsweek and Time decided to weigh their poll the way they did. This is like trying to rebuild a nuclear reactor without ever having taken Calculus. Pointless. You simply don't know WHY that weighting is there. It could be based on any number of reasons derived of historical data or something else. You're just automatically assuming that they're wrong for using that weighting but you have absolutely no factual basis for asserting that. The point is that we don't have enough information right now to resolve this conflict between the polls.

You may be right on that point, but it still doesn't matter. You nor I know why Newsweek and Time decided to weigh their poll the way they did. This is like trying to rebuild a nuclear reactor without ever having taken Calculus. Pointless. You simply don't know WHY that weighting is there. It could be based on any number of reasons derived of historical data or something else. You're just automatically assuming that they're wrong for using that weighting but you have absolutely no factual basis for asserting that. The point is that we don't have enough information right now to resolve this conflict between the polls.

Hmm. Not really sure what to say back to you here? - I think the fact is we both want GWB to win - We both know the "MO" is on our side - But with that, we still both have to stay consistent and look for the facts and truth (that is what makes us different then the Dem's.....facts matter to us) -

And with that notion - We can honestly say that when a poll shows an over-sampling of Rep's.....(that in the horse-race section of this poll.....the results are skewed somewhat) -

They are certainly not accurate - And then we can adjust their weightings to a more historically accurate weighting and see what the results of their poll would be closer too -

Thus looking at the NewsWeek poll (and CNN/Time) if you break the numbers down it appears GWB lead is around 5pts or a little more....but not near the 10pt mark - (at least not at this time) -

Lets wait until Gallup and IBD have their new polls out later next week -

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.