Stupid media poll tricks in Iowa*
This Des Moines Register Monday headline
attempts to sell what appears to be bad news for Gov. Palin's presidential prospects in a key primary state - "Iowa Poll: Sarah Palin's favorability rating slips among Iowa Republicans." But does it really? Let's investigate:
Sarah Palin's popularity has declined among the very voters the former Alaska governor would need to impress first were she to seek the 2012 Republican nomination for president, The Des Moines Register's new Iowa Poll shows.
Palin's favorability has ebbed with Iowa Republican likely voters, whose most active members make up the state's presidential caucus electorate, in the past 15 months, according to the poll taken this month.
While a solid majority of this group still views Palin positively, the intensity has waned as Palin has become a regular national media presence but done little to cultivate a support base in Iowa.
The shift is not dramatic, but it raises questions about whether Palin would have staying power in Iowa as she moves closer to a decision on a White House run, pollsters say.
[More]
So if the shift is not dramatic, why the dramatic headline? The difference is a decline of just six points in Gov. Palin's favorability among Iowa Republicans from November 2009 to this month. The casual reader of this article would assume that the Register did not measure Gov. Palin's favorability in the state between November, 2009 and February 2011, but the casual reader would be mistaken.
Not mentioned anywhere in this article is the fact that the Des Moines Register
reported June 11, 2010 that it's "latest poll" at the time showed that Sarah Palin had a favorability rating of 58 percent among likely GOP voters. So among the same group being measured, Gov. Palin's favorability has actually
increased by seven percent over the last six months. So why did the Register's story today compare her numbers to a 2009 poll instead of its own more recent 2010 survey? We can only guess, but Craig Robinson of
The Iowa Republican seems to have the Register pegged:
"It seems that with every election cycle that passes, the Des Moines Register’s bias towards Democrat candidates and liberal causes has become more and more apparent. Not only has the Register’s liberal bias more noticeable, the state’s largest newspaper acts like it doesn’t even care about how it is perceived anymore."
If the Register wants to compare Gov. Palin's numbers now with those of 15 months in the past rather than 6 months ago, it should, in the interest of fairness, at least mention somewhere in the article that it conducted a poll in 2010, and the results of that poll indicate that Gov. Palin is viewed more favorably by likely Republican voters today than she was in June of 2010. But the Des Moines Register is not at all interested in being fair to the GOP's first female vice presidential candidate. It would rather distort the facts by the sin of omission instead.
An additional caveat about the Register's latest poll: The portion of the sample that are likely Republican voters consisted of only 189 people, which should explain the astronomical margin of error among that subgroup of plus or minus 7.1 percentage points. That's more than twice the MOE of most statistically sound scientific surveys. In contrast, the 2010 poll of 501 likely Republican primary voters had a margin of error of plus or minus 4.4 percentage points, still not ideal, but a much more reliable confidence level than indicated by the MOE of the Register's latest poll, as any student who has taken STAT 101 well knows.
This story should serve as a textbook example of how the leftist media manipulates its own polling data to paint a distorted picture of how Sarah Palin is viewed by an electorate that same media long ago lost touch with. We can only conclude that the Des Moines Register is joining its liberal media fellow travelers in pushing an anti-Palin narrative, which is not surprising. But the way it goes about doing it is both disingenuous and despicable.
- JP