Read the above link for a list of questions in the USATODAY POLLS
I wonder if there are sources of bias that are not at all apparent-I've noticed in my area which is split pretty evenly between Red and Blue-there are many more Kerry bumper stickers than those for Bush. Also many conservatives have told me they feel uncomfortable to put it mildly in stating their views. I have become reluctant and discouraged if not outright intimidated in speaking up also.
Is it possible that pollsters are viewed as part of the liberal media elite and therefore Conservatives/Republicans are uncomfortable speaking frankly. Don't the polls have names like CNN,Quinnipiac and other media type or university origins to them. Who are the pollsters-are they students on summer vacartion,what do they wear? What are their own leanings? Subtle bias may creep in. What race are they?
Has anyone looked at pollster/respondee correlations?
I'm not suggesting deliberate fraud or poor design but sure would like to address these issues.
Also are there possible pro-Republican sources of bias?
Is there anyone able to address these issues?
Did a similar phenomenon happen in the 2002 elections?
If this is a source of bias-it will certainly disappear in the voting booth.
It's an interesting issue, Mike. One thing that occurs to me on reading the USA Today questions is that in any questions asking who you would vote for, Kerry or Kerry-Edwards is always mentioned first, Bush or Bush/Cheney last. Don't know if that's true for every person polled, but it might have some effect. There's also the issue that the questions encourage people to think about Bush's record as President, so they may think quite deeply about him as a candidate. But Kerry has never been President so there is no equivalent question asking people to think about his record.
In other words, the structure of the poll encourages more reflection on Bush than on Kerry in order to answer the pollster's questions. But in order to vote, citizens may think equally deeply about Kerry's record, and the GOP will no doubt encourage them to do that with advertising between now and November. Voters thinking more deeply about Kerry is unlikely to be a good thing for the Democrats.
One way of assessing the kind of bias you're asking about would be to look at opinion sources other than media polls, sources such as the Iowa Electronic Markets (link: http://www.biz.uiowa.edu/iem/) There, people "buy shares" in various outcomes - Bush wins, Kerry wins, etc. After the election, winners' shares are worth something, losers' shares are worthless, and this is real money. The idea is that players in the market are not expressing their personal preferences or opinions about who would do a good job. Rather, they're betting on who is going to win, and doing so not as a pure gamble (like a slot machine) but on the basis of knowledge. They collect knowledge from all sorts of sources - media news, talking to people at the office, etc. - and make their bets about who is going to win. The market has a good record. To the extent that the media polls are out of sync with the IEM, I would think that is an indication of bias in the polls.
Posted by: Patrick Brown | July 30, 2004 at 09:05 AM
I call the biased polls POLLAGANDA.
It is really propaganda in the poll.
The lesson is: Dont be afraid to speak your mind. Fear of expression your true views is on small step down the slippery slope of political correctness and "1984"-style conformity.
Liberating Iraq blog
Posted by: Patrick | July 31, 2004 at 02:47 AM
What the left needs more than anything else is a November surprise. Now is no time to tell the truth. Save it for 02 Nov.
Posted by: M. Simon | August 01, 2004 at 05:26 PM
It's now or never. We should start working.
Posted by: phoenix az painter | July 29, 2011 at 02:57 AM