About | Portfolio | Backup | Archives | PayPal Tip Jar | Amazon Tip Jar | Shop@Amazon
Advertising


Search BillHobbs.com
Stats, Etc.


TTLB Ecosystem Stats
Powered by FeedBurner


« Dueling Banjos | Main | Let Them Drink Sand II »

November 9, 2007

Trumpeting a Worthless Poll

The Tennessean trumpets the MTSU poll showing Fred Thompson leading Hillary Clinton by eight points in Tennessee, 50-42, and stresses that the race is "a statistical dead heat" because of the poll's four-point margin of error.

The poll is worthless, but let's pretend for a moment that it is accurate. The MTSU pollsters also stressed that the margin of error means Hillary and Fred could be tied - but if the margin of error swung the other way, it would mean that Fred leads Hillary 54-38 in Tennessee.

The poll is worthless, but as we pretend it is accurate, consider what it really says: It says that Fred Thompson is the only candidate in the Republican field who can beat Hillary Clinton in Tennessee. After all, while Fred leads Hillary by eight, she's tied in the poll with Rudy Giuliani at 43 percent. Factor in the poll's margin of error and it's possible that Hillary actually leads Giuliani 47-39. The margin of error could also mean Rudy actually leads her 47-39.

What all that tells us - if the poll wasn't worthless - is that Hillary can't crack 50 percent against either of the Republican front-runners in Tennessee. It also tells us that Fred is the only front-runner in Tennessee, of either party, who can. If the poll wasn't worthless.

But the poll IS worthless. Here's why: The MTSU students doing the poll merely did a telephone survey of 593 adults in Tennessee. They didn't limit the poll to only registered voters, much less to just "likely" voters, which good political polls do. And as Bob Krumm noted in the comments to yesterday's post at VolunteerVoters.com about the MTSU poll, the poll's data was weighted to reflect the U.S. Census Bureau's latest available projections for age, race and gender proportions within the state population.

The MTSU poll didn't weight the results for voting propensity. Says Krumm, "An 18-34 year old's opinion was identical mathematically to a 65+ selection, even though the latter group is much more likely to vote."

The MTSU poll is worthless.

Posted in Campaign Season

Comments
Post a comment
Comments Policy: Your comment is subject to deletion if it is off-topic or includes foul language or personal attack. Readers, please email me if you find comments that include egregious violations of this policy. Comments may not post immediately - do not post twice!









Remember personal info?






Email this entry to:


Your email address:


Message (optional):




back to top
Advertising

blog advertising is good for you
Video Ad Slot
To run your video ad here, contact me at bill-at-billhobbs.com
Archives
Blogroll