I’m not surprised at all to hear about bias in individual members of the media. I’m not rash enough to generalize the whole “media” based on a few reports. People tend to remember the news reports they disagree with more, anyway… but the members of the media are just as human as the rest of them. The better ones manage to keep their bias out of their work, however…
It’s amazing what you can find out when you get sick of looking for cites and just ask.
I emailed Knowledge Networks regarding their Bush speech methodology, and got a response:
Here is the substantive portion verbatim:
**
That enormous whoosh-ing sound you just heard was the air rushing out of Ned’s assertion that the 88 percent figure of support for Bush’s proposals post-speech was “grossly misleading.”
Would have been nice if they’d told me the number of people surveyed who watched and the number that didn’t. But the breakdown of political affiliation for both groups is interesting in and of itself. (And the organization stands by its claim that its survey is legitimate, scientific and has a discernable margin of error. That leads me to believe that a statistically significant number of respondents watched the speech.)
Judging by the fact that most of those who watched were Democrats, support apparently came from hearing what Bush had to say, not so much from political ideology going into the speech. That shoots to hell this Ned quote:
The guy’s comments in defense of CBS not airing the results on TV was interesting as well. Aren’t State of the Union addresses always during TV’s Sweeps Week? Aren’t they always interrupted by scores of partisan applause moments? Didn’t stop them from running snap polls before, though.
Methinks someone wants to be diplomatic and maintain a lucrative client …
Milo, you’re a dogged sort and I admire that. But the most recent debate on this subject (alluded to in your OP) should have convinced you that left-wing media bias is an illusion.
Not only is any “political” bias countered by a pro-rightist “economic bias” (remember, there are no wealthy liberals), but no bias exists at all. This has been conclusively proven by a left-wing media watchdog group. And even if there is bias, it just shows Dan Rather’s intelligence. Besides, this wouldn’t be considered bias in some European countries.
Just had to share some of my favorite arguments by opponents during the previous debate. 
Not really. The original debate was leaning towards whether or not Dan Rather was significantly affected by a liberal bias in his news reporting. The fact that the survey was legitimate, AND a proper indicator of American attitudes towards Bush’s speech, I’d say that the answer has taken a step closer to “Yes”.
::cough Alec Baldwin cough::
Probably not overall. But the possibility exists for individual members of the media to be subject to their natural human biases. In fact, I’d say it’s a given.
It shows his “intelligence” because he chooses not to report something because he dislikes the information being presented? If this is an accurate portrayal, it just shows his pettiness, and bad reporting skills.
Actually I think it demonstrates a number of things beyond the fact that I am full of crap
First is the tendency of people to assume there must be deep meaning behind every action. Neither of us has sufficient background in a network newsroom to evaluate this decision. You are assuming that an explicit decision was made to ballance Buffy (or whatever was on next) against poll results rather then a network decision that Buffy wasn’t going to be delayed.
The 88% figure remains garbage no matter how you slice it. The answer from CBS really doesn’t tell us a whole lot about the decision within the time alloted. If we go back to the beginning of the thread remember you balanced the decision to air an interview with two women against reporting the poll. Business reasons don’t tell us a thing about this decision. You are so anxious to demonstrate bias that you jumped on a silly rational when you would have been better to stick to your original premise.
What this indicates to me is that the network recognized that the results of the poll have always been fluffy. The poll has likely been done in the past to provide some context to discuss what is supposed to be an important event but is difficult to cover in a compelling way. The longer term design of the survey with a follow up poll a week later is a bit more compelling.
The fact that the Network spent a ton of money on this poll doesn’t help us much. Faced with a decision to air something that cost a million dollars or something that cost nothing but is judged (rightly or wrongly) to be more compelling there is no contest. You can always use the report somewhere else, as they did.
The breakdown of party affiliation provided in the email tells us a few interesting things first of which being that you don’t know what most is
We see that polsters regularly withhold data from us when reporting polls. This is a constant source of irritation to me. People who report polls should have some sort of requirement to make raw numbers and questions available so we can fully evaluate polls.
We see that asking party affiliation can be misleading. How did they arive at that number? Did they ask for party registration, which one voted for last election, which party one identifies with? Each of these questions would have resulted in a different set of numbers. And while we are on the subject, why do conservatives assume independants are half left wing until they want to argue Perot cost Bush Sr. the election?
In this case its entirely obvious that party affiliation is misleading. To deny this is tantamount to an admission that you don’t understand statistics. It remains absolutely clear that the viewing audience was not representative of the sample so any conclusions drawn from their opinions have limited value.
Obviously it was silly of me to guess that a sense responsibility had something to do with the initial decision. This doesn’t change the fact that the reporting provides no evidence of bias to anyone not determined to see it.
Unless you can provide a RIGOROUS analysis (no, a laundry list won’t do) that conclusively demonstrates that such decisions consistently favor the left it is meaningless by itself. To give you some context, I have a hell of a time not believing that the New York times has a right wing bias. I know you will find that ridiculous but that is the nature of personal observation.
I was referring to this debate , which went considerably far afield from Dan Rather.
I guess I set up those straw men a little too delicately. 
To clarify: I agree with Milo’s basic premise. Snapshot polls like that one, dubiously significant though they may be in the long run, are routinely reported. Rather’s ignoring it has a bad odor.
And so it goes.
My apologies. I of course meant a plurality of those who watched Bush’s speech and responded in the survey identified themselves as Democrats.
Ned:
**
Really? I think it’s looking better and better, the more I find out. I can see why you would have a stake in that assertion, however.
**
Agreed on not knowing what goes on in a network newsroom (although I’m fairly certain I know more than you. If it’s important to you to know how, email me and I’ll tell ya.).
My assertion would be that, if past practices by a network news department over many years of presidential addresses in January or February were changed in this, the first such speech by a Republican president, that may be an indicator to some of bias. Particularly given other circumstantial evidence about the way that network’s news department leans. (Which you can spend an hour or two wading through at the linked site devoted to just how biased this particular network and this particular anchor seem to be.)
Conclusive evidence? No. Compelling, or at least curious, evidence? To many.
**
First of all, the answer was from the people who conducted the poll for CBS, not from CBS.
“Silly?” I think I’ll just let it stand there and speak for itself. Several others have noticed what I notice about it.
**
I’ve already asserted that how you and I feel about polls is largely irrelevant to this debate. I’m more interested in CBS’ demonstration over many years of how they feel about polls, and how their past practices quickly changed after the Bush numbers came in.
Compelling? To any unbiased journalist, survey results showing 88 percent of a cross-section of Americans (fairly evenly split among Democrats, Republicans and Independents - with more Democrats than any other segment) who watched Bush’s speech liked the proposals they heard immediately afterward rather screams “compelling.” It rather screams, “We’ve got something here that we didn’t expect, that people aren’t going to realize unless we tell them.”
That is pretty much the definition of “news.”
Would there have been further news stories days later, showing that support lessening as people learned more? I would think so. It doesn’t make that initial snapshot any less newsworthy, however. Especially considering that news department’s past practices.
**
Oh, come on. Do you think ANY polling organization checks party registration? They ask the question, “Do you consider yourself a Republican, a Democrat or an Independent?” Handsome-looking straw man you’re building there.
**
A. It’s only marginally different than the overall segment of those surveyed; and
B. It still has enough newsworthiness to report, especially in light of CBS News’ practices when Bill Clinton gave speeches.
**
It’s not conclusive evidence of anything. But to use your own words, one could only dismiss it as readily as you do if one were “determined to see it that way.”
Please share your calculations. For mine, I am assuming that the answer was binary (approve/don’t approve), and a .05 level of significance. I get a sample size of 253 for those who watched/listened.
M.E. = 1.96 x SQRT(PxQ/n)
Based on Milossarian’s post, we know that the sample isn’t very biased at all, in fact it included 8 extra republican respondents at the expense of the independents. Take the most extreme scenario, that all of the extra republicans answered “approved” but would have been independedents who answered “don’t approve,” and the 88% statistic drops to 84% (actually 84.8%, but I rounded down).
I know this analysis still makes a lot of assumptions, but I’ll conclude with the point that no polling firm in their right mind would/could make the kind of mistakes you would need to assume to discount the results as much as some here have.