Pollster.com

Articles and Analysis

 

Is Everyone on Vacation This Week (?) "Outliers"


Andrew Gellman solves the "nonpuzzle" of close election polls.

Chris Cillizza notes an "attacking unfairly" gap in this week's CNN poll (Pollster reader Gary Kilbride caught it too).

Mark Mellman sees significant changes in Barack Obama's base of support, as compared to Kerry in 2004 and Gore in 2000.

Nate Silver sees similar patterns in the Gallup data.

Jennifer Agiesta delves deeper into how increased black turnout might have affected the 2004 outcome.

Allan and Sheri Rivlin think John McCain needs to articulate a credible economic plan.

Tom Jensen points out the online recording of an automated PPP interview call.

SurveyUSA rounds up their recent polling on the economy.

The Associated Press has questions and answers for polling skeptics.

Late update:
Kathy Frankovic says the economic issues remain dominant, as they were in 1992

 

Comments
distantobserver:

Andrew Gelman's (I believe this is the right spelling of the name...) contribution is most interesting, and I think it's well worth following the links and comments there as well.

Someone should notify Mark Mellman that it's simply bad style to draw on "a handy comparison" without providing sufficient data to take a look at the source one refers to. Apart from that his article is an interesting read.

The AP Q&A is mostly scratching on the surface, but it's obviously not directed at those who like most pollster.com regulars have given at least some thought to the problems of public opinion research.

The title of this week's outlier collection raises the most interesting question on polling in the doldrums: How much does vacation bias the results?
Unfortunately pollsters rarely, if ever, tell us about the trouble they have to get their sample size together. If I usually need to place 5,000 calls to get a sample of 1,000, I'd be tempted to conclude there might be a bias when all of a sudden 6,000 calls were needed for the same result.
We had a similar effect around Thanksgiving, then again - though to a lesser degree - around Xmas, I think, and in both cases the bias was unfavorable to Obama.
Maybe Mark can provide some information about whether the firms recognize the possible problem, and about how they - in their usual secrecy - deal with it?

____________________

Ale_Catter:

Mark,
I don’t know how you get the poll average in the national trend, but in my estimation Obama is still leading with enough margin. I think is a mistake to count every survey, you should count only the last result from every company. All Gallup tracking surveys have basically the same sample design, so you are multiplying its influence because they present the results on daily basis. The idea to average surveys is to cancel the bias that any sample design could have. So the correct thing to do could be to average the last results from different companies (or samples -Gallup/USA is different from Gallup tracking). If you do that Obama has a lead of +4.6 since 7/8, or +4.2 since 7/18, or +3.9 since 7/23, or +5.2 in the last 5 surveys (Rasmussen,Gallup, Economist, CNN,Research2000).
Best,
Alejandro (QMSS – Columbia University ; Poliarquia – Argentina)

____________________



Post a comment




Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR