feed icon

Touching It With A Ten Foot Poll

By Kleinheider Posted on July 10, 2008 at 3:45 pm

Steve Ethridge of the polling firm of Ethridge and Associates, LLC defends himself in this thread against charges that he wildly miscalled the Van Hilleary/Jim Henry GOP Gubernatorial primary in 2002:

In the poll to which you refer, the undecideds (41%) and no responses (9%) were 50%, leaving the base of those who chose a candidate to be 50%. Out of the poll, Hilleary had 30% which, when projected using the proportionate allocation method, projected to 60% for Hilleary (i.e., 30% divided by 50% who were decided). The margin of error on the poll’s re-percented based was +/- 6.9, meaning that the poll would be accurate within the margin of error if the election outcome were anywere between 53.1% (i.e., 60 minus 6.9) and 66.6% (60 plus 6.9). Hilleary won with 64.3%, well within the margin of error.

Moreover, the poll was conducted more than two weeks prior to election day, and anyone who knows anything about politics knows that a lot can change in two weeks; especially the two weeks immediately before the election. Even so, a point in fact is this: When correctly analyzed using the polling industry’s standard way of making projections, which is the proportionate allocation method, in 28 years of polling experience, Ethridge & Associates, L.L.C. has never once been outside the margin of error on poll taken within the last two weeks prior to election day.

bookmarkpermalink

Comments

2 Responses to “Touching It With A Ten Foot Poll”

  1. Sonny Parks writes
    July 10th, 2008 6:37 pm

    I am completely dumbfounded that Steve Ethridge is trying to pass this poll off as a scientific poll. If you read the small print of the press release “300 likely Republican Primary voters selected at random from people who voted in the August 2006 Republican Primary in District 7, with results weighted to the most recent, comparable Republican Primary in this District, which was the August 2004 non-gubernatorial year.” From what I can tell Steve Ethridge is weighing this entire poll on the complimentary votes Marsha Blackburn received in 2004 with no opponent. Turn out in Shelby County that year appeared to be higher than normal due to a few heated races on the ballot and it looks like the turnout was significantly lower every where else in the district. Weighing this poll on 2004 numbers puts Shelby County at around 60% of the district which is complete nonsense because anyone who read the Commercial Appeal today knows that Montgomery and Williamson County are two of the fastest growing counties in the last few years with Shelby facing a steady decrease in population. Was this accounted for in the poll? Doesn’t look like it. Was the fact that there is a heated race to replace John Wilder accounted for? Probably not. It also appears that Steve Ethridge had to make a series of negative statements about Blackburn to get people to change their vote which is not surprising now that a few people have appeared to point out that Ethridge is on Tom Leatherwood’s Finance Committee which means he has a vested interest in skewing the numbers to favor his candidate and make them appear to look better than they actually are. My guess is that this poll is going to be a repeat of the polling numbers Ethridge released the night before the Jim Henry-Van Hillary race showing Henry only down by 10 points. 24 hours later when the votes were tallied Jim Henry lost by over 30 points. Based on Ethridge’s current and past perfomances my money is betting that Leatherwood is about to face a similar outcome.

  2. Wintermute writes
    July 10th, 2008 7:09 pm

    Review this new guy’s findings during the Herenton-Chumney-Morris race, sans his rationalizations.

    Moral? Call Yacoubian.

Leave a Reply




Partly Cloudy

Recent Comments

The Collective

The Latest from NashvillePost.com

Archives