Friday, April 9, 2010

Andrew Malcolm: Sarah Palin in the Spolight

*
As Sarah Palin prepares to address the Southern Republican Leadership Conference in New Orleans this afternoon, LA Times political blogger Andrew Malcolm has an early morning analysis of a new poll which shows some improvement in her unfavorable numbers:
A new CBS News poll finds that after all the negative press, an impressive 24% (think that'll get Palin-haters going?) of its sample views her favorably.

CBS found that a mere 38% view Palin unfavorably, a whopping three-point improvement from her 41% unfavorable rating in January. By comparison, Gallup recently found Reid's unfavorable rating to be 45% and Nancy Pelosi's unfavorable rating to be 54%. So how's that healthcare change working for ya?

Palin is self-employed so doesn't have a job approval rating. But the RealClearPolitics average shows the Democrat in the White House has a 46.4 job disapproval rating and an approval average of 47.2. The same average gives Congress a job disapproval rating of 75.5, or about twice as bad as Palin's unfavorables. (Can you hear the screaming yet?)

Unlike Congress, however, Palin isn't facing any voters come Nov. 2.

Not surprisingly, Palin is viewed least favorably by liberal Democrat women in the Northeast. She's viewed most favorably by conservative Republican men in the Midwest.

One other intriguing and counter-intuitive aspect to the CBS poll:

Despite her prominence in politics and the media in recent months with her national best-selling book tour, speeches and frequent Fox News appearances, those who claimed to be undecided or not to have heard enough to have an opinion about Sarah Palin increased from 32% to 37%.

And if you believe their answers, we've got a non-existent bridge to nowhere for sale.
What we find most striking about the CBS poll is the methodology (PDF) the network employed to arrive at its final set of numbers.

Of the 858 respondents who were surveyed, the breakdown by political party was 347 independents, 271 Democrats and 240 Republicans. CBS, obviously not feeling these proportions were sufficiently biased against Gov. Palin, put its big thumbs on the scale and "weighted" its random sample to take some of the randomness out of it. The network cut the number of independents to 324 and jacked up the number of Democrats to 294.

Weighting is a scheme that more accurate pollsters than CBS -- Scott Rasmussen, for one -- do not employ. Polling firms that tend to have greater reliability also only sample likely voters. CBS doesn't do this.

- JP

2 comments:

  1. Another psychological broadside launched by SeeBS the day after the biggest conservative event in the Mid-West, the wildly enthusiastic Palin/Bachmann rally in Mpls.

    There is such a huge disconnect between polling results, and stuff you can actually see with your own eyes. Such as: sellout crowds whereever she goes, huge rallies whenever she campaigns for other candidates or appears at free events such as Searchlight and Mpls, 2.7 million copies of her book sold, one of the most sought after speakers, getting paid big bucks for an 8 episode documentary. Things like that just don't jive with the low favorable ratings all these polls show.

    ReplyDelete
  2. You can buy a random sample of people to call. Weighting a poll means you tampered with the sample.
    I trust Rasmussen even more now.

    ReplyDelete