NBC/WSJ poll: Despite defeat, Obama's jobs bill is popular

Phone polls skew to people who have time and desire to answer them. Conservatives are automatically less likely to answer them, as conservatives generally are working, spending time with family, or have a distrust of media.

During the day, who would be interested in talking to pollsters on the subject? The polling starts with the answering of the phone, and then continues with the first sentence or two. If the first sentence starts with “I would like to ask you a few questions about the President’s jobs bill” vs “I would like to ask you some questions and this may take X minutes” you get a different type of respondent.

So polls like these generally reach those who are able to answer the phone and answer questions. That lead to unemployed, retired people generally. Then depending on the framing of the first few sentences people who are highly opinionated on the topic may choose to stay on the line whereas people who don’t care will hang up. Specifically speaking of this poll, if purely by random, because you call between 9am-5pm, you got 25% of the respondents who are unemployed, they will skew the results obviously. Retirees too are perhaps more likely to favor government initiatives because they rely on SS and Medicare for example. And so on.

All that, just for starters. Then there is the wording of the questions, and other psychological aspects to polling.

With all the experience accrued in political polling, all the underlying demographic information about populations publicly available, and the available mathematical underpinnings, you would think that the science of polling methodology would have advanced to a point where there would be common standards for accuracy widely followed in the polling industry. Unfortunately, this is not the case because too many pollsters either have a political axe to grind and/or cut corners in their methods to save money.

In the particular case of the poll discussed in this thread, the article reveals next to nothing about the specifics of the methodology used. Why? I suspect that if they spelled out the specifics, their results would be discredited.

Exactly. All you really need to say is MSNBC and you know it probably won’t be scientific. But hey, it gives them something to fill their news cycle with. They are a business, and have a “customer base” that fits a certain demographic. I wouldn’t be surprised if MSNBC has a non-exclusive contract with a small polling outfit that allows them to conduct, say for example, 20 quick, rudimentary polls each year. So when they want to feed their viewing audience some red meat, or make them feel good, or give them talking points for call-in shows and office water-cooler debates they can generate some data quickly, easily, but nothing to take any more seriously than an episode of Family Guy. It’s all geared to make the customer enjoy the experience of watching the program.

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.