ndert2
Follow us on Facebook

Latest Comments

ND Outdoors Sites

FishingBuddy
NoDakOutdoors

Featured Link

Meanwhile

Login Form



Support NorthDecoder

Search This Site

Loading

Feature Stories

Personal
The Low Road

Amazon Search

JoomlaWatch 1.2.12 - Joomla Monitor and Live Stats by Matej Koval
The Fargo Forum Can't Afford These Hits To Its Credibility PDF Print E-mail
Written by Chet   

RalphsSI feel kinda bad for Patrick Springer at the Fargo Forum.  He was apparently assigned the task of trying to make "newsworthy" the polling results the Forum paid for from "Essman/Research," a firm that obviously doesn't know the first thing about political polling.  On Friday night, the Forum released poll results they claimed showed Heidi Heitkamp was behind slumlord Rick Berg by ten (10) percentage points.  Interestingly, the Forum had another story ready to go -- the same day -- about how its poll results were being mocked by anybody who knew anything about polling.

Unfortunately, I already wrote this blog post once (last night) and have since lost it.  So I'm going to be relatively brief in explaining just a few of the many fundamental flaws with the Essman/Research poll the Forum released.  

If you compare demographics from the Forum's Essman poll report from June with demographics from the Forum's Essman poll from last week, you'll see that the percentage of North Dakota's "likely voter" population that is between the ages of 18 and 30 went from 15% in June to 21.6% in October.  That's almost a 7 percentage point increase in the number of 18-30 year olds in North Dakota.  There has also been a .2 percentage point drop in the number of people 65 years old and older.  That's because the number of baby boomers going into retirement is quickly dropping.  The population percentage that is Independent increased by 5 full points during the past four months, while the percentage of Republicans dropped a statistically significant four points, from 38% down to 34.8% and Democrats similarly dropped from 23% down to 19.4%.  

(Now THAT would be news, if it were true.  A full 8% of North Dakotans went from affiliating with a party, to being "independent."  STOP THE PRESSES!!!)

Why do I point out these "changes"?  

Because they didn't happen.  None of these "changes" happened.  There are some things that a pollster (who's worth his salt) knows. A good pollster will know that in North Dakota, for example, the percentage of people who self-identify as Democratic-NPLers is probably in the 28% to 29% range.  The number of Republicans is probably in the 38% to 39% range. The percentage of voting population that's in the 65+ age category is probably in the 22% range.  The percentage that's 44 and under is probably in the 38% range.  The percentage that's in the 45 to 64 age range is probably 40%-ish.  And these percentages -- a good pollster will know -- do not swing by 5% or 10% between June and October.  A worthwhile pollster will know these things because they've had their ear to the ground in North Dakota for the past 20 or 30 years, they have their ear to the ground here, and they understand probabilities, statistics, trends and demographics.  A qualified, worthwhile pollster will, for these reasons, "weight" their poll results so that the real-world demographics match the ever-changing demographics of a random phone-dialing survey.

A knuckle dragging hack, on the other hand, will do what Essman/Research did and the Forum propagated by simply surveying the first 600 people who'll answer their phones, they'll assume that "random sample" is representative of the entire population of North Dakota as a whole -- facts and reality be damned -- and they will blast those polling results to the dumber-than-average readers of the Fargo Forum.

Now, if Patrick Springer had wanted to really inform the inforum readers, he would have told them about these unexplainable swings and the realities of survey participant demographics.  But no.  Instead he released the "raw" polling data with no explanation of its value (or lack of value). He should be embarrassed. I'm embarrassed for him.

Nobody with any credibility (or brains) thinks Rick Berg is leading in North Dakota by 10 points. Not even Rasmussen Polling, the notoriously right-wing-leaning pollster. Rasmussen was polling at nearly the exact same time as Essman, and their right-slanted results are Berg 50% and Heitkamp 45%, with a +/- 4 point margin of error. Heitkamp's pollster also released its own poll showing Heidi leads Berg by three precentage points, with a 4-point margin of error.  Here's what Heitkamp's pollster had to say about the ridiculous Essman poll:

The Essman poll has an absurdly small number of Democrats in their sample (19%). This fails to reflect the known electoral behavior of North Dakota voters. Indeed, 2008 exit polling shows Democrats making up 28% of the electorate, while in 2004 Democrats comprised 27% of the electorate. No exit poll was conducted in 2010 (nor will one be in 2012) but the likelihood that Democrats have shrunk from a consistent 27-28% of the electorate to just 19% is near zero.

Mellman Group Polling Report

The race is tight, no matter how badly Berg's spin doctors at ForumComm want Berg to be leading. 

The other thing is this: if Berg had a lead outside of the margin of error, his own pollster -- Public Opinion Strategies (POS) (who got really interested in my "Knotty Lock" fiction novel story, for some reason) -- would have released its polling results. According to the FEC's website, Berg has paid POS over $120,000 this year for polling.

The best part of the Forum's irresponsible, meaningless polling story is that the Forum has promised to continue to release more of its meaningless, unweighted, raw polling data -- including the absurd Romney (57%) to Obama (32%) numbers -- this week.  Springer will continue to embarass himself and the Forum by telling us how Stutsman county is divided on Heitkamp with 7 out of 16 polled saying they support Berg and 7 out of 16 saying they support Heitkamp, and 2 undecided.  Really Patrick?  Sixteen people in Stutsman County were asked who they support, and you -- Mr. Springer -- are going to project that across the whole county in your news story?  Sure; you say the polling sample isn't statistically significant, but do you tell your readers just how incredibly statistically insignificant those numbers are?  If it's statistically insignificant (and it is), why do you even write about it?  

Sixteen people?  You could just as well walk into the Pizza Ranch in Jamestown, ask everyone there for lunch who they support, and release those numbers as being "newsy" numbers.  

I'm surprised people still read the Fargo Forum. They really blew it in June when -- in polling on the Heitkamp vs. Berg race -- they filtered out anybody who didn't know how they were going to vote on the ballot measures.  They aren't making things any better this week by releasing meaningless poll results from an apparently incompetent political pollster.


Comments (6)add comment

nimrod's leftist cousin said:

...
Fargo Yellow Rag.
 
October 24, 2012
Votes: +0

Star said:

...
New poll shows Berg with a 5 point lead. And Cramer leading Gulleson. I am really losing hope with the people of ND. It just seems that all you need is the letter (R) behind your name and they will vote for you even though you have repeatedly not served them well. It is well known that Berg has voted to privitize Social Security, & followed the Tea-Baggers and right-wingers in Washington in regards to the Ryan budget. Yet women, and older, retired people continue to support these knot-heads. Cramer takes campaign contributions for the very people he is supposed to regulate and protect us and the beautiful landscape and nature of ND, and he gets away with it saying that they gave him the money because of the way he thinks! Could it not be that they gave him the money to shape his thinking? Sad, SAD!
 
October 24, 2012
Votes: +0

What the Heck said:

Hoeven
His last minute endorsement of berg is undoubtedly making a difference. Too bad ND voters just don't get it. I am shocked at how many of my friends on Facebook are republicans. While I am independent myself I definitely lean left on social issues. That said I am having a harder time casting my vote for Obama. If the republicans had a better candidate I would probably vote for them. On state issues and representation it is a no brainer. I am voting for every dem on the ballot. The republican reign of corruption in this state must be stopped.
 
October 24, 2012
Votes: -1

Marty said:

...
The question is whether Hoeven will regret getting caught on tape praising Berg's character and integrity, after Chet's book comes out.
 
October 24, 2012
Votes: +0

What the Heck said:

Agree Marty.
You can almost sense Hoeven's apprehension in the ad. Although it could be the fact he has never taken a political risk his entire career. His last endorsement for Goettle had to seem relatively low risk and look how that turned out. Personally I cringe whenever I hear any politician pounding their chest about integrity. It should be apparent by actions alone. Two of a kind, IMO.
 
October 24, 2012
Votes: -1

Guest said:

Math
When you say "full 8 percent" you mean 6.8%? And I disagree with the headline, the Forum had no credibility to lose.
 
October 29, 2012
Votes: +1

Write comment

security image
Write the displayed characters


busy