Monday, October 20, 2014

Looking to November: The KHOU/Houston Public Media Poll

Last night, on their 10 PM newscast, KHOU released the results of a poll jointly sponsored by Houston Public Media.

Abbott, Patrick command big leads in KHOU, Houston Public Media Poll. Doug Miller, KHOU

Full poll results can be found here including a breakdown between Harris County and Statewide.

In short, the poll has Abbott with a 15 point advantage over Wendy! Davis and Dan Patrick leading Leticia Van de Putte by 12 percentage points. 

Looking over the poll questions (and I'm assuming the questions on the results page are the only ones asked, it's fairly bare-bones focusing just on the races in question and some key issues.

One conclusion by Bob Stein, Houston's bicycling authority, and go-to guy for political analysis for the TLSPM,  puzzled me however:

(from the KHOU story linked above)
The timing of the survey happened to coincide with a period before and after the Davis campaign launched its hard-hitting "wheelchair ad" criticizing Abbott for opposing plaintiffs suing for damages while Abbott himself collected a multi-million dollar settlement over the 1984 accident that left him a paraplegic. As a result, the poll offers a clue about whether the controversial political spot influenced voters.

"It's had no impact on the distribution of the vote," Stein said. "And it's had even less impact on mobilizing people who otherwise, we think, would be supportive of Wendy Davis."
Based on the questions asked, I'm unclear with how Stein is coming up with that conclusion. There is nothing in the poll that addresses the ad, nor is there anything that might suggest the ad's impact. In short, you can't make any inferences on the ad's effectiveness in any area by looking at this poll.

What we're left with is Mr. Stein's personal opinion on how the ad has been perceived, and that's hardly an opinion coming from a disinterested source. These are issues that I have with most Stein polls, not that his methodology is flawed (it's impossible to tell without the cross-tabs and he's not the best at releasing those) but that the fuzzy nature of the responses allow him to make sweeping generalizations regarding results that aren't fully addressed in the mechanics of the poll.

In short, he's editorializing. Injecting his opinion into the poll results because that's the way he thinks things are going. Or, maybe there's something not being released, a question relating to the ad perhaps, that provide some data behind his ruminations?

Who knows?