Letters to the Editor
-
It didn't have to be this way.
One reason we're in the sorry state we're in is that with the impeachment of Clinton, the Democrats failed to constitute a Committee for Revenge and appoint me chairman. The chief goal of the CFR would have been the total elimination of the Republican party. This did not happen. I apologize for any inconvenience this may have caused anyone.
-
@Mike Sulzer
But that is an example of a fluctuation that does not average out, one with a strong deterministic effect. That is very different from random fluctuations that do tend to average out.
I meant it was an example of a fluctuation that was short (sorry, I should be more specific). No, it wouldn't average out, but its effects would be minimized by it being short w/resp to the window width. The general problem is that you don't have a way to suspend time, nor a true baseline from which to measure. Some of the compensations are truly more complex than just time span, by the structure of the survey you can try to estimate the effect of specific events and you can difference. I don't personally know all the tricks, I just listen to the resident demographer explain these things to me, but I do understand the basic framework.
Which means that the results only have real significance with reference to these hypotheses and models. And that is not the way the results are usually used in the media.
It isn't quite that bad! Yes, your statement is accurate, but the models tend to be close, as are the hypotheses, and there are reasons for the formulation of the questions, the stratifications, the codebook, and all that. Many of these carry over from survey to survey, or are the accepted formats for particular kinds of surveys. For instance, there is a reason why at least 1 or 2 of the questions on presidential approval surveys are always worded the same way, there is a reason why the sampling time is usually a constant for public opinion surveys, etc.
You are absolutely right that all of this isn't reported, for a lot of surveys, especially academic ones, it is available in general form. These NEP jobs are special in the level to which they proprietarize (?) the data.
You would also be right if you criticized the media comparing quickie polls, standard opinion polls, exit polls, and votes as all the same and mutually comparable with no theoretical argument. Likewise epidemiological surveys or whatever.
The thing shooter referenced was a political criticism of an adaptation of an epidemiological method for estimating deaths due to war (the Lancet survey). The method and adaptation are fairly standard, but then beyond that everything gets weird, for instance, criticizing med/bio/epi people for not publishing their raw data is absurd. It's a jealously guarded quantity in a world where most people use census data off tapes, and people would only publish it if they felt there were no more papers to write from it. Repeatability ought to mean if you doubt it, go get your own money belt and $20K and do another one, but not in the critical article, where a Harvard prof acted like a Johns Hopkins prof is always going to hand over the data, and then pulled survey problems out of a hat (e.g. curbstoning) and made it seem as if he had a reason to think they were problematical.
There is an oblique mention of the fact that there is a hypothesis in the explanation of the statistics, I did hear that news media did mention that in 20 surveys one would be outside the margin of error on average -- a clear reference to the hypothesis involved in the confidence interval for the survey.
I don't think it's realistic in today's climate to think that they are going to explain all the intricacies of the survey methodology to everyone. Not that it wouldn't be a good thing to give at least a synopsis, but very few people seem to recognize that survey design is a whole discipline (usually demography, sociology, or epidemiology, sometimes operations research).
-
@Yellow Dog - re: "Partisanship versus Bias"
Not only is the American style of "objective, above it all, omniscient" reporting and commentary dishonest; it's dull.
The British press is interesting; the writing is good.
American commentary is predictable; it has the feel, almost always, of something you'd expect to get from a technician running a word processor, not a writer.
The narrow paradigm, where the horse race and soap opera meet, and where issues are only occasionally addressed (and when they are, only superficially), gives voters very little to work with.
From election cycle to election cycle, it's a feedback loop. It's devolved into a "Let 'em eat cake" form of political journalism.
It's insulting, when you throw in the low melodrama of broadcast news. Chris Matthews, with what he said last night, may have unwittingly undermined the basic foundation of the whole enterprise. It's a statement, which, if it was studied thoroughly acted upon by serious media reformers, might put Matthews and his ilk out of work.
As a sampling of intelligent, literate opinion, writers on this blog are saying they're leaving MSM reporting in droves.
The dogs here don't like the dog food.
The product is harming democracy, not advancing or promoting it.
-
Jebbie...
While you're on this 'roll', please grab your crayon and whip off a note to the NYT about Junior, and MSNBC about Tweety.
Who knows. We might get lucky.
"I'll take it under advisement." ;~)
FYI... Zuckerman has still been there on occasion...
Crayon?
-
Tweety
I just heard Chris Matthews say that John Edwards "looks cute."
It's getting to the point where I watch just to catch him uttering idiotic/weird/sexist/latently gay Freudian slips.
-
@Reilly - I was so taken by your post that I didn't read on
How very well put and agreeable and civil and educated and just plain fine!
(on another topic)
Regarding Chris Mathews' negatives to Hillary and positives to Rudolph, I think he is probably quite unconscious of his patriarchal misogyny. He is a son of the patriarchy, and has by osmosis (as it were) absorbed its tenets.
How do I know this? I too am such a son. The sound of Hillary's attempts to add gravitas to her voice in a speech are grating (may I say somewhat comic) to me. My educated conscious self is working on overcoming this, but it is deep. (Like racism, an ape-mind subtext).
If she were more true to her own person as a woman, expressed, she would do better with me, anyway.
(Rudolph is a fascist fraud, no bones about it).