feed icon

Explaining The Asterisk

Posted on July 3, 2008 at 2:34 pm

Ed Cromer of the Tennessee Journal shines a bit of extra light on the recent polling on the U.S. Senate race released by the Bob Tuke campaign:

In his “Tuesdays with Tuke” e-mail this week, Nashville attorney Tuke offered only sketchy details of an “800 sample” Global Strategies Group poll but said it showed him 30 points ahead of Padgett with 57% undecided. Wednesday, the campaign said there had been a typographical error. The corrected survey results showed Tuke with 32%, Padgett 8%, others 4%, and 57% undecided — which because of rounding add up to 101%.

The correction included an asterisk note that the results were from a sub-sample of Democratic primary voters — meaning the margin of error is substantially higher than if the entire 800-person sample had been used. This suggests that the full sample was used to gauge Tuke’s standing versus Alexander. Those results weren’t given.

Comments

One Responses to “Explaining The Asterisk”

  1. JWelch writes
    July 3rd, 2008 9:30 pm

    and the fact that the full results weren’t given suggests that Lamar’s numbers were at least as good in Tuke’s poll as they were in the rasmussen poll a month or so ago. The question is, “were they better?”

Leave a Reply




Recent Comments

The Collective

The Latest from NashvillePost.com

Archives