The Rasmussen Reports is an electronic publishing private public opinion polling outfit. A couple months ago the firm released the results of a poll asking Americans about Iran. The survey of 1,000 registered voters was conducted following release of a government report saying that Iran halted its nuclear weapons development program in 2003. Not surprisingly, we have a very varied opinion as to the veracity of that report. In answers to specific questions:
- 52% see Iran as an enemy
- 4% say Iran is an ally
- 39% view Iran as somewhere in between the two
- 18% believe that Iran has halted its nuclear weapons program
- 66% say Iran has not stopped its nuclear weapons program
- 21% of men believe Iran has stopped the weapons development
- 16% of women believe Iran has stopped the weapons development
- 77% say Iran's nuclear program is for weapons development
- 67% believe Iran remains a threat to our national security
- 19% disagree
- 14% are not sure
- 59% believe the United States should maintain sanctions against Iran
- 20% disagree and 21% are not sure
- 47% and 34% believe it is Very Likely that Iran will/is likely to develop nuclear
- 29% of liberal voters and 8% of conservatives believe Iran has stopped its weapons program
- 56% believe Iran should be stop nuclear weapons development prior to any high level meetings with the U.S
- 36% say U.S. relations with Iran will get worse over the coming year
- 33% say they will get better
- 11% of U.S. voters think America should apologize to Iran for "crimes" against the Islamic country - one of the prerequisites demanded by the Iranian president before he will agree to meet with President Barack Obama
- 100% of me thinks Ahmadinejad and (at least) 11% of those surveyed has a screw loose
The margin of sampling error for the survey is +/- 3 percentage points with a 95% level of confidence.
~~
Posts
5 Comments:
Some questions/concerns/opinions:
1) How were the "registered voters" chosen?
2) What are the responders' ages, political party, education level, religion, working or not?
3) I submit that if each question on said poll didn't have the choices "I don't know" and/ or "I don't care" the results are invalid.
4)"100% of me thinks Ahmadinejad and (at least) 11% of those surveyed has a screw loose"
You can count me in on that one.
An example of using statistics to tell lies?
Like JEG indicates, accurate statistical sampling is dependent on many factors. However, non-leading questions asked of a general enough base should produce a reasonable estimate of opinion. This poll, like most others, reveals that some of those sampled are obviously ignorant, misinformed, or supplying smartass responses... otherwise how can we we explain the 4% who say Iran is a ally?
otherwise how can we we explain the 4% who say Iran is a ally?
5th column? They thought it was Iraq, or maybe they thought it was Israel. Maybe this was an attempt to locate spies from Iran. Maybe these are immigrants from Iran, who may not support the current government in Iran, but could never imagine their homeland as an enemy of a country they chose to move to.
There are probably some Texans, who moved from Colorado, but do not consider Colorado to be the enemy of Texas. Wait until everybody is fighting over stimulus funds. It will be every beggar for himself. :-)
Polls are not known for their objectivity in questions. The people writing the questions have an agenda. Maybe it is that they are trying to find out about a product/candidate that they are considering selling/supporting. They want to find out what problems they might have, before they spend a lot of money on selling/supporting.
Maybe they are using the poll to try to shape opinion. If successful, maybe next poll will show 5% calling Iran an ally, or maybe only 3%.
A lot is in the question and the way it is worded. a lot is in the data and the way it is presented.
3) I submit that if each question on said poll didn't have the choices "I don't know" and/ or "I don't care" the results are invalid.
I agree. People should not be forced to choose an answer if they don't know or don't care.
Many polls are also skewed. Here is an example of a skewed poll.
1. Really agree.
2. Moderately agree.
3. Slightly agree.
4. Disagree.
Properly constructed answers should have a neutral answer "I dont know" and an equal number of "agree" and "disagree" answers, like this:
1. Strongly agree.
2. Slightly agree.
3. Don't know/neutral.
4. Slightly disagree.
5. Strongly disagree.
Post a Comment