They've actually done studies that show word order in questions does indeed matter when it comes time for people to register their opinions. I'm not saying that has anything to do with the results of the poll (as I think Fox News is a bane on the television industry... not that I watch TV much anymore, but when I did, it was), just something to think about.
Also, the percentages don't have to add up to 100%. They're asking of, say, the 100% of people who watch CBS, how many people believe there's a link between Saddam and al-Qaida. 56% of the people who watch CBS believe that. I mean, we could all sit around and estimate how many people that 67%/56%/whatever% really did believe that we found a link, and then determine if that's "most," but we all have better things to do. (I hope.)
If anything, the 16% of people who watch PBS/NPR that believe there's a clear link between Saddam and terror surprise me the most.... or at least the difference between them and Fox news/CBS.
This did occur to me, but that's not what stated in the terminology above the poll. It says "Primary news source for those who believe: Since the war ended, the U.S. has found Iraqi weapons of mass destruction." This clearly implies that, for example with Fox, they are saying that that 33% of those who believe weapons of mass destruction have been found in Iraq use Fox News as their primary news source. If what you say is true, what it SHOULD say is that for all respondents who say that Fox News is their primary news source, 33% of those believe that they have found weapons of mass destruction in Iraq. So you are making an inference from the data which is probably accurate, but it's not how the text that accompanies the graphic presents the data (it's quite possible that whomever created the graphic didn't understand the survey).
But let's assume you are correct (as it does make more sense given the %'s shown), then doesn't it become important to know how the "Primary News Source" data breaks down? We know that 3334 adults responded total, but how many responded with Fox News as their primary news source? How many responded with PBS/NPR? What if the number of Fox News viewers included in the survey is only 3 (or of PBS, ABC, etc.)? Now if I was doing the type of survey that you suggest they were doing, I would have found an equal number of respondents for each network. That way you can have an apples/apples comparison between each news source. But how was it actually done, you can't tell. Maybe I'll try to find where this thing actually comes from, maybe there is more info.
Finally, don't give me that shit about "most". The word DOES have a meaning. And it's pretty easy to tell when you have an either/or case, it's whatever the majority is. Assuming your interpretation of the study is correct, this would mean that MOST people who use Fox as a news source believe that Saddam was working closely with Al Qaeda, MOST people who use CBS as a news source believe that Saddam was working closely with Al Qaeda, and MOST people who use anything else as a news source do not believe that Saddam was working closely with Al Qaeda. It's not that hard.
Everything that is wrong in this world can be blamed on Freddie Prinze Jr.
*sigh* Obviously, I failed to explain myself clearly.
ges, you managed to rebut my point about "most" excellently - and then proceeded to say what I was saying in the first place! I am fully aware that above 50% constitutes a majority. What I was saying (and I think I was making fun of everyone by this, but obviously this didn't seem to get across, either) is that of the 3,334 people, we could figure out how many watched Fox, how many watched CBS, etc., etc., then multiply that number by the percentages that they responded (with the primary news source) to figure out if over 50% of people total believed there was a link between Saddam and Al-Qaida. Evidently, I just borrowed that phrasing from the people who wrote the survey.
Also, no, we don't know how many people responded with each news source - which, once again, was my point.
By the way, this looks like another hysterical and exaggerated thread title. This is, perhaps, becoming a pattern amongst threads started by Canadians (and Grimis of course). I will try to keep a closer watch.