Pollster.com

Articles and Analysis

 

POLL: Newsweek Approvals


A new Newsweek national survey (story, results; released 6/22) finds:

  • 26% approve of the job George Bush is doing as president, "a record low" for Newsweek; 65% disapprove.

  • 25% approve of the way Congress is handling its job; 63% disapprove

** The release ommitted methodological details. We will fill those in once they are available.

 

Comments
Watcher:

Uh-oh, must be an outlier, right guys?

____________________

Nope-- not an outlier. See the quickie note that's been up at PoliticalArithmetik since shortly after the Newsweek poll arrived.

The trend estimate is 29.9. That puts Newsweek 3.9 points below trend, but the confidence interval is approximately +/- 5% around the trend. With the trend at 29.9, estimates between about 24.9 and 34.9 would not be outliers.

Of course the trend estimate changes with each new poll, so what constitutes an outlier depends on where the trend stands at the time of the poll (and possibly later if new polls show either more agreement or more disagreement with the particular poll.)

As always, we run the same analysis of each new poll.

____________________

Nope-- not an outlier. See the quickie note that's been up at PoliticalArithmetik since shortly after the Newsweek poll arrived.

The trend estimate is 29.9. That puts Newsweek 3.9 points below trend, but the confidence interval is approximately +/- 5% around the trend. With the trend at 29.9, estimates between about 24.9 and 34.9 would not be outliers.

Of course the trend estimate changes with each new poll, so what constitutes an outlier depends on where the trend stands at the time of the poll (and possibly later if new polls show either more agreement or more disagreement with the particular poll.)

As always, we run the same analysis of each new poll.

____________________

Till Eulenspiegel:

I'd love to see the liberal/conservative and Dem/Rep/Ind breakdown of the Congressional approval numbers, if available.

My expectation is that they're continuing to lose ground among liberals and Democrats, after giving up the fight on Iraq.

____________________

Till Eulenspiegel:

Ah, I see the second part of my request is available the MSNBC site.

Fascinating that the percent approving is almost identical in all three groups.

____________________

Bill Mitchell:

The last Newsweek Survey oversampled Democrats by 17 POINTS. (not kidding)

Princeton Survey Research Associates who conducted the poll are just shills for the left. This time they were smart enough not to release their methododoloy since they got so much grief for it last time.

If you want a more accurate number, look to Rasmussen who has Bush at 35%.

This Newsweek poll isn't scientifc, it's agenda-based propaganda.

____________________

Bill Mitchell:

The last Newsweek Survey oversampled Democrats by 17 POINTS. (not kidding)

Princeton Survey Research Associates who conducted the poll are just shills for the left. This time they were smart enough not to release their methodoloy since they got so much grief for it last time.

If you want a more accurate number, look to Rasmussen who has Bush at 35%.

This Newsweek poll isn't scientifc, it's agenda-based propaganda.

____________________

John:

Rassmussen always has him eight or nine points above everyone else, check their record. Princeton is an entirely respectable polling organization as is Rassmussen. They just have different methodology. This is the third, I think poll to have Bush well below 30%. Given the hot water he's in with his base over immigration, it's not a stretch to believe he's lost a bunch of them who have been keeping him in the low thirties for months.

____________________

Till Eulenspiegel:

I hope by "oversampled by 17 points" you don't mean "17 points away from 50/50". You are aware that there are significantly more registered Democrats than Republicans in the country, right? You are aware that pollsters weight according to various demographic data to adjust for over/undersampling, right?

____________________

Bill Mitchell:

By 17 points I mean they had 17 points more Democratic Leaners in the sample than Republicans. Hope you can do that math.

In 2004, exit polling showed Dems 37%, Reps 37& and Ind's 26%.

Are you trying to tell me that Dems have picked up 17 points on Republicans in 2 years? Yes, more Democrat registered voters than Republicans but:

1. Nowhere near 17 points more, and
2. Most of them don;t vote - that's why we had a Republican Congress for over a decade and a Republican President for the last 6 years.

And PULEASE don't try to justify Princeton's bogus sample by saying it's just like all the other MSM polls that ALSO wildly oversample Democrats.

Every election season, Rasmussen comes closest to reality. It is because he weights his sample fairly, not this bogus 17 point stuff.

____________________

Bill Mitchell:

I believe that the SAMPLE used in Princeton's last BOGUS poll was:

Democrat Leaners 51%
Republican Leaners 34%

Sorry, but that kind of sample is just rediculous. It is not scientific. It is agenda-based propaganda. Period.

____________________

Anonymous:

Partisan self-identification does change and has changed, guys. You don't have to check a box to do that. People are no longer identifying with an unpopular party. 500 R's and 500 D's would result in a terribly incorrect poll because it's an inaccurate reflection of the country.

____________________

Chantal:

Bill, I'm not sure "shills for the left" would show Congress with a 25% approval.

____________________

Newsweek/Princeton Survey Research Associates are among the best pollsters in terms of clear disclosure of the sample's properties. The May survey Bill is talking about includes the party distribution at

http://www.msnbc.msn.com/id/18491981/site/newsweek/

The distribution of party id is

22% Rep
36% Ind
35% Dem
7% None, Other, Don't Know

When the Independents are asked if the lean towards one party or another, the distribution INCLUDING these leaners is

34% Rep
51% Dem

which are the numbers Bill reports above.

Most of the reports of partisanship are based on unleaned samples, but some do report with leaners. This produces more confusion about what the distribution "really" is.

More importantly, as you filter the sample from Adults, to Registered Voters to Likely Voters to Actual (Exit poll) voters, you see the gap between Reps and Dems narrow. That gives a range of values that people pick from.

I wrote a piece about this distribution here

http://politicalarithmetik.blogspot.com/2007/05/partisan-variation-in-2007-polling.html

For a now somewhat out of date look at partisan trends over 2005-first quarter of 2006 see

http://politicalarithmetik.blogspot.com/2006/03/partisanship-moves.html

____________________



Post a comment




Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR