Sunday, April 10, 2005

Russ Steele finds our halos unholy

updated Apr 14

In a comment on our Friday post, local conservative blogger Russ Steele begs to differ with its content, and does so at his weblog.

Russ is making two specific objections:

First, he says we do need "balance" in reporting climate science issues.

(Russ, is it just the specific case of climate science reporting that you disagree on, or all kinds of science reporting?
Do you agree that in the case of alien abduction experiences (which Mooney explains do have a scientific, non-alien explanation), "balanced reporting" does a disservice to the readers?
What's your overall (i.e., not attacking a minor point) impression of the April Scientific American editorial on this problem, "Okay, We Give Up"?)

In support of the need for balance, Russ cites the 'contrarian' pair (environmental economist Ross McKitrick and mining executive Steven McIntyre) critiqued here (see also Real Climate's myth #4.)

Will we get down into the point-by-point climate science trenches with Russ? No. NCFocus has no expertise in climate science, nor a belief that developing one would be the best use of our sadly limited mental resources (and time): outsourcing judgment to the experts is likely to yield far better results. Don't take it up here, Russ; take it up with the scientists who devote their lives to studying climate science and aren't being funded by those with a vested interest in a particular outcome.

Also see Krugman's April 5 column - "Thirty years ago, attacks on science came mostly from the left; these days, they come overwhelmingly from the right, and have the backing of leading Republicans..."

Second, Russ attacks Rogers' claim (here(PDF), on the PACNC website - don't forget to link to what you're criticising! It shows you have faith in your position and in your readers) that CRI's "82% support parental notification" automated phone survey results boiled down to only 5.4% of actual survey responders saying yes - Russ says the %response was % of total calls made, not of people who answered any survey questions.
([Apr 14] sorry, this confusingly worded to the point that I'm afraid to clarify it for fear of changing the meaning and making it wrong - maybe another night, when I'm awake...)

(another apparent real flaw: she does also state that 200 responses are needed for statistical validity, which sounds (to NCFocus) more like someone's rule of thumb than statistical truth; doesn't the # needed for statistical significance depend on the results?)
Apr 14 update: By "validity", she's probably referring to statistical power, not statistical significance as we originally assumed.
Also this Boston Globe article is relevant, in an amusing way - them Easterners got mighty high standards:
"If it were our poll, we would be very upset if it were used in a way that did not make it clear that there were that many [20%; CRI's was ~ 90%] nonrespondents," said Francis J. Connolly, senior analyst for the Kiley and Co. public opinion research firm.
...
Also amusing: Rodgers is troubled by the editor's influence over both the editorial and news pages. "As senior news editor, Cliff Schechtman makes the final decisions," he said. "He also actively participates in the paper's editorial policy..."

Russ says:
Typical survey responses are 2 to 3 percent. This survey’s response was over 6 percent, a valid survey response. It was not a bogus survey, as suggested on the PACNC web site.
Another point though - it's only "valid" if the responses of those who did answer the question are representative of the public as a whole. It seems likely that the few who don't hang up on automated phone spam are likely to be substantially different from the general population.
In any case, it's our considered opinion that the PACNC Q&A and Myths and Facts have info that's more relevant to what the policy should be.

(Russ does link to CRI's response to McCall's April 5 column in the Union, which we hadn't done, so here it is.)

4 comments:

Russ Steele said...

Anna:

"funded by those with a vested interest in a particular outcome."

Please do not fall into the trap of attacking the scientist, when their science challenges cherished points of view. If you can prove their science wrong with data, OK. Just attaching the person and the funding shows how weak your argument is. Data is true regardless of sides, regardless of where the funding comes from, government, environmentalist, or industry. Data is data. It does not have emotions.

Russ

Anna said...

short answer: "consider the source" is a pretty good heuristic, when you don't have the time to look further.

longer answer:

Russ, the April Fools editorial in Scientific American wasn't just a joke, it was making a point. For the record, can you a) verify that you understand what point they were making, and b) state whether or not you agree with it?

Another question - since you delve into the climate science, you look at evidence from all sides before coming to conclusions, right?

If so - presumably you read Real Climate regularly, to get the climatologists' perspective?

In the unlikely event that you missed it, here's their takedown of the McIntyre and McKitrick "random noise" claim.
[Fairly complex statistical explanation, then] "Only one of the parties involved has (1) had their claims fail scientific peer-review, (2) produced a reconstruction that is completely at odds with all other existing estimates...and (3) been established to have made egregious elementary errors in other published work that render the work thoroughly invalid..."

Would it be reasonable to ask that you check RealClimate before repeating claims that they [claim to] refute, and - if after doing so you still want to forge ahead - also address their points in your post?

You don't have to do this, of course - your blog is yours to use as you wish - but it's what scientists do.

Anna said...
This comment has been removed by a blog administrator.
Anna said...

(just to clarify, lest you think you missed something really good - my above "removed by the author" comment was just a test comment, to see if Blogger was behaving, since someone had reported it wasn't.)