In other words, when measuring the efficacy and accuracy of a news agency, you can’t ignore the basic cognitive abilities of the general audience.

For example, if the average IQ of Fox’s general audience was 45 points less than the IQ of NPR’s general audience (hypothetically speaking of course) they would, naturally, perform poorly at correctly remembering any facts presented when compared to NPR’s audience.

NPR tends to attract a more educated audience (in my opinion… feel free to argue that point) so if it’s not at the top of the list with a reasonable margin it must be performing poorly in comparison.

In short, if you’re going to compare the accuracy and efficacy of a news service’s audience based on the accuracy of their audience’s recall, you must measure the cognitive abilities of the base audiences in general. You can’t assume the average NPR listener will have the same cognitive abilities as the average Network Morning Show listener.

Furthermore, this is just one issue in making such a comparison and coming to any conclusion about the news service in general. Another issue that comes to mind is audience attention level. For example, many people may have a morning news show on but may give it far less attention than those listening to an evening broadcast.

The real fallacy here is assuming causality. Surveys are fairly good at showing correlation, but are very poor at showing causality. It shows that there is a significant deviation in factual recall between listeners of The Daily Show and network morning shows, but it doesn’t show why. Jumping to any conclusions about the accuracy or efficacy of these news services is fallacious and misleading.

1) The principal investigator in the original PIPA/Knowledge Networks study was Steven Kull of PIPA. Knowledge Networks fielded the survey but the KN employee most directly involved was Stefan Subias.
2) The original PIPA study deal with misconceptions about the Iraq war and showed that the principal news source for the respondents was highly correlated with the number of misperceptions. For example, viewers of Fox News were much more likely to believe that Iraq had been involved in 911 than listeners of NPR (a misconception that still persists mind you!). The more recent Pew work cited in your post that was drawn on by journalism.com and The Atlantic was targeted at more general knowledge. For example: “Who is the president of Russia?” The answers to these questions are very straightforward and not as subject to ideological bias. So people who consume a lot of news are going to get these questions right more often. Intelligence and education will also correlate highly with correct answers.

If you re-ran the earlier PIPA work with current questions such as “Is Obama a muslim?” or “Is China drilling for oil off the coast of Florida?” I surmise that one would again uncover that viewers of Fox News would be more likely to get these questions wrong than viewers of The Daily Show or NPR listeners.

I’ll have to dig into this more recent Pew study as well. Knowledge/awareness of current events isn’t the same as being aware of the facts. There’s no way that O’Reilly viewers are nearly as *correctly* informed as NPR or Daily Show viewers!