|
Issues/News>
The DNews Poll Interpretation
March 23, 2008
SLTrib.com 2/23/08
Out of Context :
The Tribune's political writers' blog.
Sunday, March 23, 2008
Poll vault
It's been said that good political polling is equal parts
art and science. But the same goes for interpreting poll
numbers.
For an illustration of that maxim, and a look at the art
and science both falling short, take a look at Sunday's
survey in the Deseret Morning News. The story says that
incumbent Rep. Chris Cannon could face a tough re-election
bid. That part is pretty much common sense. But for proof,
the DNews looks to a Dan Jones poll that shows 30 percent
favor Cannon to 19 percent for his top Republican
challenger, David Leavitt, and less for Jason Chaffetz.
These numbers, however, are basically worthless for a
couple of reasons.
The sample consists of "residents." That's not voting-aged
residents or registered voters or likely voters or
registered Republicans. It's just residents.
The public preference doesn't mean squat at this point. The
general public -- those sampled in this poll -- won't get a
say until November. Before then there will be delegate
selection, a convention and probably a primary. Right now,
it's all about the delegates, and they don't get picked
until Tuesday.
That, of course, doesn't stop the candidates from offering
meaningless spin to the meaningless results, including
Leavitt's odd assertion that he feels good about the
numbers since he's only been campaigning for a month. You
could see where that might be encouraging -- if it were
true. But Leavitt filed his "Statement of Organization"
declaring himself a candidate with the Federal Elections
Commission in May of 2007 -- approaching a year ago -- and
he had spent $132,593 just through the end of 2007. That's
a lot of campaign money to be spent not campaigning.
Wonder how much more he's spent in the last month, you
know, since he started campaigning.
-- Robert Gehrke
(Tip of the day: Jason's last name is pronounced "Chay-fits")
|
