November 11, 2006

Weekend Question 2: How did the Columbus Dispatch Polls Fare Against Actual Results?

ANSWER: For the third time in a row, very poorly.


Three times in a row — as was the case last in November 2005 and May 2006, the Dispatch’s vaunted mail-in poll differed from actual results by double digits in almost all cases, and all errors were in the conservative/GOP direction:

Governor –
Predicted Strickland Margin: 36
Actual: 23.5
Error: 12.5 points
Direction: Conservative

US Senate –
Predicted Brown Margin: 24
Actual: 12
Error: 12 points
Direction: Conservative

Attorney General –
Predicted Dann Margin: 24
Actual: 5
Error: 19 points
Direction: Conservative

Secretary of State –
Predicted Brunner Margin: 21
Actual: 14
Error: 7 points
Direction: Conservative

Auditor –
Predicted Sykes Margin: 10
Actual Taylor Victory Margin: 2
Error: 12 points
Direction: Conservative

Treasurer –
Predicted Cordray Margin: 28
Actual: 15
Error: 13 points
Direction: Conservative

If the Dispatch can’t do any better than this, it should either ditch the mail-in poll, or do a top-to-bottom redesign. If it continues to do its polling as usual, reasonable people will have to conclude that they’re doing it to influence election results, and not to honestly attempt to portray reality.


Selected Previous Posts:
- Nov. 5 — Dispatch Polls Are STILL Within Their Track Record Margin of Error
- Oct. 17 — Why “The Polls” Consistently Get It Wrong
- Sept. 29 — The Polls Done by That Columbus Newspaper Need to Be Dispatched to the Trash
- Nov. 9, 2005 — Worse Than Worthless — Ohio’s Polls



  1. Yup,

    Fair to say the Dispatch poll sucks at this point.


    Comment by Jerid — November 12, 2006 @ 2:13 am

  2. [...] Bizzyblog shoots fish in a barrel. Fun but a tad unsportsmanlike. Turns out the Poll was wrong by anywhere from 12 to 19 points in the liberal direction. [...]

    Pingback by » Bizzy Takes Apart the Dispatch Poll — November 12, 2006 @ 9:05 pm

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.