New Year's Resolutions - Poll Design
Disclaimer: I don't know why we Americans make New Year's Resolutions. When I asked my personal guru, all I got was this lame article. So, if anyone can tell me where, when, and why this tradition was started I would really appreciate it.
There's more.
Resolutions often focus on personal goals, like losing weight, or going to the gym more often. But this is an opportunity to highlight the importance of professional goals in the new year. For example, Rick Wagoner (Chairman of the board for GMC) should resolve to make a profit in 2009, and Bernie Madoff should think about how to make an honest profit. And even though pollsters did much better in 2008 than they did in the last two presidential election cycles, there are still plenty of ways to improve polling quality. For example: Today I was aimlessly surfing the Internet (don't tell my boss) and stumbled upon this blog entry from April, 2008.
This delightful little blog post, written by a libertarian who dabbles in social policy analysis, could be billed as a How NOT To Guide for pollsters. If you have been reading Student Redux's series, "A Plebian's Guide to Polls", you will see how the RNC broke most of the rules and best practices in our series. I'm sure the DNC has been guilty of equally egregious acts of polling stupidity, but the ideology of the poll is not my main interest. I am more interested in the fact that this poll created completely meaningless data. For example:
4. Which of the following factors do you feel is most adversely affecting the economy in your area?
- Budensome Taxes
- Unstable Real Estate Market
- Threat of Terrorism
- Severe Government Regulations
- Growth of Government Spending
- Unpredictable Fluctuating Fuel Prices
To say this is a leading question would be the understatement of the year. In a way, this entire poll looks like a ham-fisted attempt at push-polling or parse polling, but asking the Republican base if "Burdensome Taxes" are "adversely affecting the economy" is -- literally -- preaching to the choir. Parse polling on these issues would have made since in April of 2008 (when the author blogged about the poll). But, most parse polls are more focused than this wide-ranging shotgun blast of ideology. It actually seems like the author(s) are so blinded by their ideology that they can't see past it.
Here's my suggestion: The author(s) of this poll should resolve to write polls with reasonably neutral language in 2009. The author(s) should also resolve to only work for Republican candidates, I rather enjoy winning.
--pluribus














what a perfectresume work!
what a perfectresume work! great job!
Not as incompetent as it seems
Reading through Berggren's post, note that he buried the lede--"After completing the survey, I was taken to a page requesting a donation to the Republican National Committee." I doubt that the data from this "survey" was used at all--it was just a scam to get gullible Republicans thinking that their opinion was valuable. Once they were psychically invested and engaged with the RNC, they would be more likely to donate. There's a world of difference between this, which is just a fundraising trick, and a real poll.
You beat me to it
Damn it, I was going to say that. Particularly noxious about it is the illusion that the RNC cares about what the people say. "Now that we've listened to you, will you help us act on your agenda?"
Disgusting.
Admin
I can't get that worked up about it
Not least because the DNC and lefty groups do it too (no cite right now, but I've personally received mail pieces at about this level of subtlety). Let's face it, if you place a high value on honest and forthright debate, you're probably not writing direct mail.