As my readers know, I have taken an interest in polling, primarily because it seems so many polls are agenda driven.
In a post last week, I pointed to a PPP poll on Wisconsin which used sample data which underweighted Walker voters from 2010 and overweighted union households. I also have focused in the past on health care polling and other pollling used for political purposes.
I’m pleased to announce that next week Legal Insurrection will have a special guest series of posts on pollling, written by Matthew Knee, a Ph.D. candidate at Yale (don’t hold it against him) specializing in campaigns and elections, ethnic voting patterns, public opinion, and quantitative and experimental approaches to political science.
The posts will run starting Monday, and will focus on polling issues which arise in the political context, such as weighting of samples, framing and sequencing of questions, and other factors.
Matt also will evaluate various public polling on the issue of public employee unions, as that issue has been in the news quite a bit and the various polling results are being used by each side for political advantage.
I hope you will enjoy the series.
——————————————–
Follow me on Twitter, Facebook, and YouTube
Visit the Legal Insurrection Shop on CafePress!
Donations tax deductible
to the full extent allowed by law.
Comments
Don't know if you've learned about it as yet, Prof Jacobson, but Col Brendan Doherty of the RI State Police tendered his resignation to Lincoln Chafee, because of Chafee's compromising his authority and refusal to listen to Doherty on the issues of the need for enforcement of laws that protect us. Chafee's eliminating e-verify, and his facilitating various cities imposing sanctuary city status.
Hey, a political scientist! Sounds right up my alley, LOL! 🙂
"the various polling results are being used by each side for political advantage."
I dare you to find a liberal site that agrees instead of just saying 'Faux Newz Polls lie!!' I say this because I was reading some coverage at ThinkProgress (I like to at least give the other side a chance) and the articles/comments there made me sick.
Really looking forward to this. Welcome fellow UCLA Bruin!
I look forward to it. I hope Mr. Knee discusses the fundamental weakness of all "scientific" opinion polling: the unverifiability of the relationship between a respondent's answers and what the respondent actually thinks. In most polls not only all the questions but all the *answers* are supplied by the pollster; the respondent only chooses which words are to be put in his or her mouth. Opinions are elicited that the respondent may not actually hold strongly enough to be worth mentioning. Thus opinion polls not only misrepresent the distribution of opinions, they misrepresent whether the opinions even exist. Except in instances of concrete preference polling, the results of opinion polls are factoids down to the core. They are not news, i.e. reports of something that happened. Polls are means of manufacturing stories that journalists want to tell instead of what newsmakers actually did. They aren't fit to print.
Looking forward to the series which should get linked everywhere! When pollsters are not intentionally manipulating or statistically innumerate, they still suffer from the same malady that sociologists and scientists do in their studies- that often it's more about them than it is about what they're researching.
You could follow up the week with an opinion poll on opinion polls and interpret the data differently through numbers sleight-of-hand. Or, do several polls framed to yield different targeted results from the same population of Insurrection readers.
I wonder if you could also cover the way the unemployment numbers are generated. Because it seems to me that it's very similar to the polling issue. A bunch of locations or people can be cherry picked to arrive at some startling conclusions. For instance, last month after declaring that we lost jobs again, unemployment went… down. There were many articles written on the incredulity of this conclusion from what we could all see was bad employment news.
This month unemployment went down even more than… last month! How?
Something stinks here and I am suspecting we won't see the revised numbers until August or Christmas, when no one is looking.
Very much looking forward to this series. I'm deeply distrustful about any polls, but particularly any funded by the state-run media. Polls, as we should all know by now, are mostly designed to shape public opinion and not reflect it. The recent PPP disgrace really typifies what polling has become.
Van Halen, if you were listening to Rush or Levin, you would know that the 0bama regime jiggered the employment figures so that the entirety of available jobs is now a smaller number. This results in a lower unemployment figure. Furthermore, they don't count those who have stopped receiving unemployment and dropped out – last I heard, 14 million jobs and workers have been "dropped" from the unemployment calculus.
Presumably by early fall '12 just in time for the election, the official unemployment figure will be 8% or lower (the supposed top number with which an incumbent president can win election) while in actuality MANY more will be unemployed with many more permanently lost American jobs.
If they can get all of us out of work for over 2 years, unemployment will drop to close to zero. They just stop counting us. Think about it.
Since I can't comment on Hot Air (registration for comments has been closed forever on there, and apparently will be till doomsday) I thought I'd comment here on a post by AllahPundit (sigh) regarding a Rasmussen poll (on-topic!) supposedly showing Governor Walkers approval rating at 46%.
Rasmussen, out of all the pollsters, seems to be one of the fairest. However, that does not mean he can't be wrong and that he might not have a Pro-Union bias. Even if his methodology proves to have been fair and he is attempting to be accurate, that does not mean his poll is proof that Wisconsin is pro-public union and anti-Walker or that the majority side with the unions. Polls are not gospel, they are subjective and only provide a sample at best. So let's not all panic over any poll, whether Gallup, Rasmussen or the much less creditable PPP.
Speaking of AP:
When it was pointed out how AP has an uncanny ability to seize on negative polls, present them as undeniable proof that the public is against conservatives and cry that the sky is falling he retorted to the effect that people were just upset at not "hearing what they want to hear." Assuming if course that Mr. AP always accepts and embraces things he doesn't want to hear, that's not the issue. The issue is that he makes a big deal out of everything slightly negative (or allegedly so) and cries like it's the end of the world. It's tiresome and people are sick of his passive-aggressive moderate antics like this and his attempts to sabotage conservatives (which he is not, despite whatever he claims) by frightening them away from their principles and positions based on thin evidence. He dismisses what other when they say that he doesn't like, but we're supposed to 100% accept his arguments. Bull.
FZ, AP has lost a lot of credibility with conservatives for exactly what you describe. I rarely read him anymore and used to follow him avidly. (And their comment registration policy is so restrictive I read HA less and less. Reg. has been closed for YEARS.)
Professor, excuse our dip in another blog pool 😉 I for one greatly appreciate your comments policy by comparison with some!