Rasmussen daily graph for 12/8/07 - Clinton up 2, Obama unchanged, Edwards down 1

[DavidDvorkin]

Long run, Obama and Edwards stuck, Clinton still strikingly down but no longer falling. Jeez.








(Pretend that the 10 and 20 lines are actually 0. Does anyone know how to make Excel label different parts of the axis differently?)

Rasmussen links:
Data in tabular form
Discussion

[Horse with no Name]

1. Rasmussen leans right.

Them having Hillary on top is par.

[DemocratSinceBirth]

3. Huh

He has HRC with an eleven point lead... As I demonstated in my previous post , aggregate polling shows Hillary with a near twenty point lead...

[durrrty libby]

5. So all pollsters "lean" right ?

http://www.realclearpolitics.com/polls/


:eyes:

[hayu_lol]

6. Thanks David...these really do help.

A lot easier to read the graphs than it is to read through a pile of polls.

[DemocratSinceBirth]

7. As An Empiricist I'd Like A Plausible Explanation Of Why Rasmussen's Numbers Differ So Greatly From

The Numbers Of Others

[DemocratSinceBirth]

2. I Have A Question

How do you explain the discrepancy between Rasmussen's results and the results of (several) other polls of roughly the same vintage... I can't believe a double digit difference can all be attributed to a tighter screen, ergo:

http://www.realclearpolitics.com/epolls/2008/president/democratic_primaries.html

[DavidDvorkin]

8. I have no idea

I've said before, when posting these graphs, that I chose to do this with Rasmussen because that's the only poll that publishes daily numbers. If others did, I'd graph them in some fashion -- or maybe Pollingreport.com would be doing it for us.

However, there's also this factor: trends, if any. I'm assuming that whatever bias a particular poll has is consistent throughout a given election cycle. If so, then whatever trends show up in these graphs are real, whether or not the numerical spread between the different candidates is. It was the hope to see some kind of trend or pattern that motivated me in the first place.

[DemocratSinceBirth]

11. Not Your Fault But It Still Doesn't Answer My Question

You are suggesting that even if a daily poll has been wrong all along we can still see trends... But how can you explain the fact that four pollsters are measuring the same thing at the same time and coming up with such disparate results....

[DavidDvorkin]

12. My guess is how they massage the numbers

My understanding is that pollsters gather the raw numbers and then calculate the numbers we see based on how they think the parts of their samples represent equivalent parts of the entire population. A difference in that calculation would produce a consistent difference between pollsters.

Not to mention how they select their samples in the first place. A consistent bias could result from either or both factors.

[bluerum]

4. and I raise you a mormon, a baptist fundy, a washed up ex NYC mayor,

,two middle eastern nations and billions of dollars of tax breaks for oil corpos and weapons manufacturers.

[DavidDvorkin]

9. I'd graph them, too

But I despise them all and I don't want to pollute my computer.

I'll vote for whoever wins the Dem nomination, and I'll vote against whoever wins the GOP one.

[flpoljunkie]

10. Rasmussen latest Favorable/Unfavorable/ rating for Hillary Clinton is 44%/55%. Obama is 50%/44%.

http://www.rasmussenreports.com/public_content/politics/favorables/election_2008_democratic_candidates_running_in_2008_presidential_election

Edwards'latest favorable/unfavorable rating is 48% to 44%.

[DavidDvorkin]

13. That's interesting

And I've noticed that those numbers seem to published more lately.

Has anyone shown any kind of correlation between those numbers and the poll numbers? You'd think it would show up eventually.