Does the company just toss out the results to the public, free of charge? Unlikely.
Or, when a poll is commissioned by, say NBC and the Wall Street Journal, what do the buyers get out of it? These are big time Republican companies in today’s world where cheating,spinning and outright lying are standard fare.
So, I’m wondering if something funny’s going on with some of the pollsters and their clients.
I can’t offer any cites, but since Truman beat Dewey, have you ever seen such fucked up polling as we have had in this presidential campaign, from the primaries onward?
So you ask, what would NBC/WSJ have to gain by skewing the polls? I don’t know., Perhaps the intent could be to keep the country thinking the race between Obama and McCain is a lot closer than it really is, so that the electronic voting machine results in November - McCain Wins!!! -are somewhat believable and the populace doesn’t go berzerk?
I have worked with other pollsters before, but not the big guys like Gallup. But they certainly take lots of business clients for marketing research, etc.
News.
I have no clue what on earth you’re talking about. So, yes, I’ve seen way more f-ed up polling. With a couple of outliers (like the unbelievable Newsweek poll a few weeks ago showing Obama with a huge national lead) that will invariably happen, I think the bulk of polling has been pretty consistent.
I have no clue what on earth you’re talking about. So, yes, I’ve seen way more f-ed up polling. With a couple of outliers (like the unbelievable Newsweek poll a few weeks ago showing Obama with a huge national lead) that will invariably happen, I think the bulk of polling has been pretty consistent.
That’s the sort of thing I’m talking about - the weird polling in this political run.
Pollsters seemed to have refined their techniques over the past 60 years so when a Gallup or other biggie pollsters came out with a poll, you could rely on its accuracy. Not these days, however.
(Pretty good answer, Ravenman, by someone who had no clue of what I was talking about.)
I assume they sell off a lot of the details. Gallup saying that it’s 48/40 is one thing. Finding out how many middle income white Catholic women aged 25-40 want to vote for McCain is another. I wouldn’t pay to find out but there’s obviously people who would.
Also, both Rasmussen and Gallup have subscription sections to their sites where you can browse data archives, expanded reports, etc. Again, not enough to entice me but I’m not a political consultant, either.
Outliers don’t mean that polling is being manipulated. They happen because sometimes pollsters make mistakes, like phrasing questions poorly or putting them in a bad order.
Again, outliers happen. Since I accept that there will be some polls that will have different results than most, I truly and honestly have no clue why you’re making the assertion that this year is worse than others.
My guess is you’re suffering from a case of confirmation bias.
Maybe, but I have a memory of (but like I said, cannot cite) several recent polls that proved way off, and various commentators including Bob Shiefer expressing surprised by it all.
I’m not an expert in polling, but I do know a bit of Sampling Theory, and the sample taken by the pollsters includes assumptions as to the makeup of the complete population. If that is changing, you can get inaccurate results. For example, presidential polls try give more weight to those most likely to vote. If this year lots of people who never voted before do, like in the primaries, certain groups will be undercounted. I’m sure the polls are trying to take this into account, but no one knows what the turnout is actually going to be, so that is a big area of uncertainty.
As for the OP’s questionl, I knew someone who was a big exec at Gallup, and he seemed to do a lot of market research. I assume the networks pay for the polls, but having “Gallup” be the generic term for polling surely is good for sales.
ETA: How dare DSeid start a thread on this very subject!
Spare me. The only thing these companies have going for them is their credibility. Rather than attribute the results to some conspiracy, wouldn’t it be easier to examine the methodology of the polling? It’s hard to claim bias when Gallup uses standard, time-tested methods of gathering data.
Also keep in mind that they deal with probabilities, nothing more. People have been getting different results for an awfully long time for a variety of reasons. There’s no reason to accuse pollsters of shilling for others without demonstrable proof of collusion.
Subscribers to polling services get the poll internals, which include demographic breakdowns of who replied, explanations of any weighting the polling company did, etc. So more than just the top line results that get reported everywhere.
Also, the really high profile polls (like for Presidential races) that are basically available to everyone basically serve as loss-leaders for Gallup and other big polling firms. Every newscast that cites the poll says the word “Gallup”, so that they’re the natural choice when someone wants some research done that isn’t of general interest.
Which is why conspiracy theories regarding presidential campaign polls being rigged don’t make a lot of sense. The whole point of doing them for Gallup et. al. is to show off how accurate they are.
I had to take a Gallup poll for a job interview one time. It was a personality/“character” type survey. I’m quite certain the hiring company paid for those services.
And the news networks still get paid for broadcasting facts, right? Why is it so hard to believe they would pay for unbiased poll numbers in which they have a financial interest? But Gallup giving the Presidential race numbers away free as a form of advertising for their other, paid services makes sense, too.
This is the main thing. Gallup does a lot of market research and consulting outside of the election polling for which the layman recognizes their name; being widely publicized for their election polls is essentially PR for their other services.
And, this. It’d be a horrible business choice, and it would destroy their credibility with their paying clients.
