Another question

[vraiblonde]

Anyone who's familiar with polling, can you help me? I'd like to know how the President's approval ratings can drop or rise by as much as 10 points in a week.

Do people change their minds that quickly? I know it's just a sampling but, even if nothing interesting has happened during that week, you might still see a dramatic change in the polling numbers.

 

[Otter]

A small sampling of people can produce large swings in polls. The polls here sample anywhere between 700 and 1100 people. Its never the same people polled on a weekly basis(I think) and it doesn't take many to produce a 5 to 10 pt difference.

 

[SamSpade]

Originally posted by vraiblonde
Anyone who's familiar with polling, can you help me? I'd like to know how the President's approval ratings can drop or rise by as much as 10 points in a week.

Do people change their minds that quickly? I know it's just a sampling but, even if nothing interesting has happened during that week, you might still see a dramatic change in the polling numbers.

Samples done correctly should reflect the universe from which they were taken with a measurable degree of error.

But depending on *which* poll you see, the differences can seem large, but they are usually small within the same poll. Different polls do not ask the same question.

 

[vraiblonde]

I was just wondering if I should start disregarding polls or if we're really a nation of flakes that changes their opinion from day to day.

I know that online polls are pretty predictable. CNN online polls tend to favor the liberal POV, Vote.com polls tend to favor the conservatives. But dirt world polls, I thought, were random and not directed at a specific political affiliation?

 

[SamSpade]

Originally posted by *archimedes*
Have you ever been polled? Why not?

A few times. One poll I participated in, it was clear they were really trying to skew the results. They would read a short intro to an issue, and then ask if I was in favor or against. If my choice wasn't along the "expected" lines, they would read some other piece of news, and re-ask, "In light of this information, do you still think A?". I noticed they *never* used this tactic when I gave them the 'favorable' answer. Once I realized they were stacking the answers, I told them I wasn't interested in continuing. [NOTE: It was a *very* liberal poll].

Another I took was a poll dealing with the arrest many years ago of several car salespersons in PG County who for whatever reason had chosen NOT to report the purchase of cars with large quantities of cash to the authorities. All of the salesclerks who did this were black, and all of their supervisors were not. The pollster ended up **ARGUING** with me about what was right, and what wasn't right. After 45 minutes, it was clear that he also wanted to skew his results to his opinion.

I've done the very simple ones - for the week of blah-blah-blah, do you think the President is doing a good job, or a bad job? NOT is he a good President or a bad President. NOT would you vote form him or against him. Sometimes they ask the negative - is he doing a BAD job, or a good one? Pollsters know that some people will always give Bush good grades, and some will always vote against him. Hence the middle of the road voters whose minds change a lot.

 

[Doc]

Originally posted by *archimedes*
Have you ever been polled?

Yes. And I always feel it is my duty to lie, or at best to give completely random answers. It's my goal to be part of that ±4% error.

 

[Larry Gude]

A more benign...

...reason may be that 35% of us pretty much disagree with anything a GOP president does and 35% pretty much agree with everything he does.

So, we have that great middle, some 30%, who like alot of what a GOP president does, but not everything and like alot of what a Democratic President does but not everything and that makes individual, specific issues very powerful because that is all it takes to tilt these folks one way or another.

In this case, the middle is reacting positively to two things they see as good: Recent economic news and visiting the troops in Iraq.

The right leaning 35% haven't moved, they knew the economy was improving and already support the war.

The left leaning haven't moved, looking for the bad, no new jobs and 'sneaking' in to see the troops as campaign ploy.