AND ANOTHER POLLING OUTRAGE – AT 8:08 P.M. ET:  Am I paranoid, or are liberal news outlets commissioning polls that they know will go in their direction? 

Added to the useless poll described in our 7:45 post is a new one released by the Washington Post and ABC News.  What does it tell us?

...the new poll shows that the political standing of former Republican vice presidential nominee Sarah Palin, who was the keynote speaker last week at the first National Tea Party Convention, has deteriorated significantly.

Huh?  This does not match other measures that we've seen, and it certainly doesn't match what we feel, something rather important in politics.  Palin's stock has risen considerably, and she's even getting praise from liberal columnists like David Broder for her recent appearances.

Although Palin is a tea party favorite, her potential as a presidential hopeful takes a severe hit in the survey. Fifty-five percent of Americans have unfavorable views of her, while the percentage holding favorable views has dipped to 37, a new low in Post-ABC polling.

And get this one:

Palin has lost ground among conservative Republicans, who would be crucial to her hopes if she seeks the party's presidential nomination in 2012. Forty-five percent of conservatives now consider her as qualified for the presidency, down sharply from 66 percent who said so last fall.

That finding should have set off red lights in the polling office.  It just doesn't make sense.

Ah, but why does it not make sense?  Once again, the last line of the story reveals all:

The margin of sampling error for the for the full poll of 1,004 randomly selected adults is plus or minus three percentage points.

Randomly selected adults.  Again.

Why would anyone take a political poll among randomly selected adults?  The purpose of political polls is to predict election trends.  Once again, the real polls are taken among likely voters.

Ignore this poll, unless the results are affirmed by another, more carefully selected sample.

February 11, 2010