Quote from: jhkim;226363And I'm saying that the analogy is broken.  Keeping a set of books for a company is a series of repetitive tasks of simple annotation.

See that...right there...that's a huge leap.

When I ran my small business, it was, for the most part, like you describe. Most of the transactions and such were pretty repetitive - once I figured out how to classify X in accounting terms, I could do Y and Z. When A came down the pike, I had to rethink some things, but it all generally followed the same process.

I'm guessing it's a bit more complicated in a publicly-traded, multinational, multi-billion dollar business. They make assessments and classifications that would make my head spin so it fits or that bizarre (byzantine!) IRS/Sarbox/whatever rule.

Quote from: jhkim;226363Scientific research isn't like that.  For example, when I was doing my PhD thesis, I spent months tuning a multi-parameter fit to my data.  I couldn't just jot down a bunch of standard steps that will let an outsider exactly reproduce what I did.

I do not think science is like my little business - at least not the science we're talking about in this venue. I'm not sure because I'm not that kind of scientist (even if my degree has "science" in it!); but I could see likening my college chem experiments to my small business accounting, and your PhD experiments in Physics to running that huge corporate behemoth.

Quote from: jhkim;226363Realize that what you're complaining about has nothing to do with global warming -- you're claiming that the whole scientific process of the past hundred years and more is broken.  The process is that if you want to verify something, you do a bunch of independent experiments and see if they all get the same result.  I think the burden should be to show that our current scientific process is broken.

Let's be clear about a couple of things. First, I don't complain about global warming. I think it would serve everyone better if we discussed the inevitable impacts and how we should shift as a society to account for them. Instead, we get into arguments about how the world is or is not going to end on December 31st, 2015.

Second, I'm not calling into question the scientific method in total. That's an excluded middle argument. The question that I keep coming back to is if we're talking about changing the way we do things to the extent that some like the Kyoto treaty would require, wouldn't it be good to hold that particular science underpinning the reason to participate in that treaty to the highest possible standard - better than even the IRS holds business where we're just talking about investors' money?

Quote from: jhkim;226363John Morrow's claim seems to be that people generally don't have any trust in science because science is a big failure.  It seems to me that the exact opposite is true.  Our scientific advances of the past hundred years are an amazing success story overall, and most people have great respect for science as an institution.

Oh, well, perhaps we're just interpreting Mr. Morrow's position differently. I will say that the distrust I see in science from people around me comes from the "It's bad for you!" today, "It's good for you!" tomorrow syndrome.

Then they see someone hailed as brilliant and influential because he predicts global warming will increase hurricane in number and intensity - and all the media with bated breath gears up for this catastrophic hurricane season.....that never comes. And then the formerly-hailed-as-influential-and-brilliant comes out and says "Well, we're not as sure, we're looking into it."

I give him all the credit in the world for having the guts to, in the face of the possible fall-out, basically admitting that the models didn't work right. But lay people see that, and liken it right back to the aforementioned not-good-today/good-tomorrow syndrome, shrug, and fill up their gas tank.

It might be the media, it might be the certainty of the proclamations of the not-good that seem to always be overturned by "further studies," it's likely a combination of the two. But that's the stuff that's eroding the layman's confidence in science, IMHO.

Quote from: jhkim;226363And I'm saying that the analogy is broken.  Keeping a set of books for a company is a series of repetitive tasks of simple annotation.  Scientific research isn't like that.  For example, when I was doing my PhD thesis, I spent months tuning a multi-parameter fit to my data.  I couldn't just jot down a bunch of standard steps that will let an outsider exactly reproduce what I did.

How does a person (even yourself, is someone asks you ten years from now what you did) determine whether what you did was correct or not?  Do you still have all of the original data that went in to your thesis somewhere?

Quote from: jhkim;226363Realize that what you're complaining about has nothing to do with global warming -- you're claiming that the whole scientific process of the past hundred years and more is broken.  The process is that if you want to verify something, you do a bunch of independent experiments and see if they all get the same result.  I think the burden should be to show that our current scientific process is broken.

Yes, it does go beyond global warming but my point is that accuracy is more important when major treaties and government initiatives are going to be based on that science.  And please note that title of the New York Times article I posted a link to earlier is For Science's Gatekeepers, a Credibility Gap and it discusses general problems with the peer review process.  

Quote from: jhkim;226363John Morrow's claim seems to be that people generally don't have any trust in science because science is a big failure.  It seems to me that the exact opposite is true.  Our scientific advances of the past hundred years are an amazing success story overall, and most people have great respect for science as an institution.

No, I don't think science is a failure.  My claim is that people are less willing to blindly trust science than they once were because it's had some big failures because those failures undermine the willingness of people to trust scientists simply because they are experts or authorities.

Quote from: John Morrow;226487How does a person (even yourself, is someone asks you ten years from now what you did) determine whether what you did was correct or not?  Do you still have all of the original data that went in to your thesis somewhere?

This may sound repetitive, but whether I was correct or not is determined by independent verification.  Actually, my thesis is better described as being itself an independent verification.  My thesis was a measurement of Alpha_S, the constant that determines the strength of the strong nuclear force.  

That constant has been measured many times prior to my thesis.  Compared to other measurements, my measurement had smaller theoretical errors from QCD (Quantum Chromodynamics), but greater statistical error than the leading measurements.  So the total error was no better than the leading measurements, but it was a check on the QCD calculations.  

Moreover, those measurement of Alpha_S will be used in future calculations.  This is how science advances.  We don't have to keep going back to what happened in 1988 (when my thesis data was first collected) to know how physics works.  

Quote from: James J Skach;226409I'm not calling into question the scientific method in total. That's an excluded middle argument. The question that I keep coming back to is if we're talking about changing the way we do things to the extent that some like the Kyoto treaty would require, wouldn't it be good to hold that particular science underpinning the reason to participate in that treaty to the highest possible standard - better than even the IRS holds business where we're just talking about investors' money?

This presumes that what is being suggested is really a higher standard.  To me, a higher standard implies better results.  Let's consider the two approaches:

1) The traditional approach, including peer review of articles by other specialists in the field and independent testing to confirm results.  Some raw data is often accessible, but is not generally posted to the general public but rather is available to others in the field upon request.  For example, my thesis included selected data from several experiments.  Specialists spend the vast majority of their time focused on their original research.  

2) The transparent approach, where scientists are required to document each step of their work such that an outsider without specialized knowledge can reproduce their exact results for each paper.  If an outsider cannot reproduce their result, then the scientist will be called upon to do this.  Some mistakes are caught, but since each experiment takes considerably longer to produce and review, there is less independent verification.  

I am saying that I don't think that #2 is inherently better science or a higher standard.  Outside reviewers are not useless -- there are some mistakes they can catch.  However, they will also make a lot of noise that detracts from the science, and it would be a big effort to bring them in -- effort that could have been spent on other ways of verification.

Quote from: James J Skach;226409It might be the media, it might be the certainty of the proclamations of the not-good that seem to always be overturned by "further studies," it's likely a combination of the two. But that's the stuff that's eroding the layman's confidence in science, IMHO.

Here is where I can't disagree with you.  

For example, the 'global dimming/cooling' everyone points to smugly was entirely an artifact of the media, and not what the scientific community was talking about as a whole.

I would posit that the current administration's anti-intellectual eight year war on science has a bit to do with the erosion as well.

Additionally, while I do have a keen interest in quantum mechanics/physics/chromodynamics, I will more or less guarantee that if I were to start reading John Kim's thesis mentioned above, I would be sawing logs in minutes.  Well, probably 30-45mins, but you get my point.  How much more the average citizen with a drastically lower interest in science?

Of course, that doesn't mean it is the responsibility of science to have high level theoretical experiments and papers accessible by the average non-scientist.  But it would be helpful in order to alleviate the confusion about science.  There are some good science writers out there, but for every Brian Greene languishing on the Science shelf at Barnes and Noble, there are five Michael Crichton's on the best seller rack at Wal-Mart.

Quote from: StormBringer;226556Of course, that doesn't mean it is the responsibility of science to have high level theoretical experiments and papers accessible by the average non-scientist.  But it would be helpful in order to alleviate the confusion about science.

Oh, agreed completely.  

I do think that science needs to make much greater inroads to being accessible to outsiders.  For example, that Creationism is still an issue at all is due not to any cunning maneuvering by the Creationists, but rather because scientists are doing badly at outreach.  It doesn't help that some proselytizing atheists (including many scientists) promote science as being opposed to religion.  

In particular, science education does tend to make claims from authority and also tends to be revisionist in its history of science, often to promote that idea of authority and great minds that make discoveries simply by being smart.  But that's a problem in explaining what science is about, not in how the process itself works.  The process itself does have problems, but they're subtle ones that often vary field to field.  Better statistics training is always good, say.

Quote from: jhkim;226501That constant has been measured many times prior to my thesis.  Compared to other measurements, my measurement had smaller theoretical errors from QCD (Quantum Chromodynamics), but greater statistical error than the leading measurements.  So the total error was no better than the leading measurements, but it was a check on the QCD calculations.

Suppose the findings of your thesis had more radical or important implications.  Suppose, for example, your measurement had been significantly different than the previous measurements and you claimed that your measurement invalidated previous measurements.  What kind of validation would other physicists want to do before accepting your revision?  Or suppose you were using the conclusions of your thesis to argue that the Large Hadron Collider and other high-energy collision colliders should be legally prohibited from being used because they pose a real and immediate danger ot the planet.  What kind of validation would other scientists want before accepting your conclusions and bringing high-energy collider research to a halt?

Quote from: jhkim;226501I am saying that I don't think that #2 is inherently better science or a higher standard.  Outside reviewers are not useless -- there are some mistakes they can catch.  However, they will also make a lot of noise that detracts from the science, and it would be a big effort to bring them in -- effort that could have been spent on other ways of verification.

While the distraction objection is legitimate, process and documentation are always a big effort and a distraction, whether we are talking about scientific research, software development, or business.  In software development, written specifications, code reviews, and comprehensive QA all take a substantial amount of time and effort to do.  In business, financial reporting, tax filings, disclosure, and environmental requirements all require a substantial amount of time and effort to comply with.  I would guess few people like to be audited or reviewed or like dealing with red tape and documentation.  So why is it ever done?  Why is comprehensive documentation and auditing an important part of ISO9000 certification and Sarbanes Oxley and so on?

Whether the red tape and documentation is worth it or not will depend on how important it is that things are done right.  If it's a research paper that confirms the conclusions of earlier work, a piece of internal software that's not critical to the business, or something like a garage sale then auditing, disclosure, and documentation likely aren't worth the effort.  But if your software or business is large, important, or incurs a high cost for being wrong, then documentation and auditing are considered critical despite the fact that it's annoying, time consuming, creates noise, and distracts from the real work that people want to be doing.  Why should science be exempted?