A ridiculous method

A recent article on slate.com by Michael Levi discusses a paper that ostensibly establishes scientific credibility based on statistical analysis. Levi's article can be found at http://www.slate.com/id/2258088/. The paper, one published in the Proceedings of the National Academy of Sciences, can be found at http://www.pnas.org/content/early/2010/06/04/1003187107.full.pdf+html. The paper uses the number of articles written, number of citations etc as criteria to argue that climate skeptics are inferior scientists compared to climate change non-skeptics. And according to Levi's article, the White House tweeted a statement saying that this article proves that " climate skeptics " aren't " cream of crop ".


Make no mistake. I believe in anthropogenically caused climate change. I believe in it strongly, almost irrevocably. But what this article describes is exactly the kind of thing we shouldn't be wasting time over. I tend towards the view that the number of publications can't be a criterion for " cream of crop " or whatever. An example of how the political elite is spending more time on playing the cheap blame game instead of solving our problems.If these scientists are fringe lunatics with no credibility, they couldn't possibly be the reason climate policy change was delayed, right ? And if they weren't important, the blame for the delay should lie elsewhere, right ? And the White House is achieving what exactly by using this questionable study to brand scientists ?

The publications of these scientists are out there, and the content of those publications can be criticized by the non-skeptics. This study is a huge waste of time.And in the future, if there is a scientific controversy, the credibility of the arguments should be judged how exactly ? How should it be judged 1 year after the controversy starts, and how should it be judged 5 years after the controversy starts ? Based on the number of papers ?

And who exactly from the White House stands behind the study which says that particular scientists aren't up to par ? The president himself ? The cabinet ? Is this method going to be standard from here on out when there is a new scientific controversy ? Is the Congress going to pass a censure motion against these scientists based purely on the number of papers they have published ? And whether the White House is or isn't up to par will be judged how exactly ? Employment numbers ? GDP growth ? Success in social welfare ? Or some looney statistical method that compares it to, say, the Johnson administration ?

There are other problems with this method. How can a group of scientists be branded as inferior ? Shouldn't this same analysis be carried out then at the level of individual scientists ? Should the career level of the scientist matter when you use this statistical analysis ? As you can see, this method is rife with contradictions. In fact, the National Academy of Sciences should take a hard look at this article, and the authors, who I think have revealed their utter ignorance of the way science works by writing this article, and pass either a vote of endorsement or a vote of censure about it, since the article has implications for all fields of scientific enquiry.

Albert Einstein was one of the few voices in the first half of the twentieth century against the probabilistic interpretation of quantum mechanics. The Einstein-Podowlsky-Rosen paradox is famous as an argument that was forwarded as an objection to the Copenhagen interpretation of quantum mechanics. The vast majority of the physics establishment believed and still believes in the probabilistic interpretation of quantum mechanics. If the method sketched out by these PNAS article writers is applied to this debate, then Einstein may well need to be branded as a substandard thinker on quantum mechanics. Many think he was wrong about quantum mechanics, but few, if any, will brand him as a substandard thinker about quantum mechanics.

Another example comes to mind. The Big Bang versus the Steady State theory of cosmology was a hallowed debate of physics till it was settled in favor of the Big Bang theorists. I don't know the paper number distribution of these two groups of scientists, but the debate was settled based on logic and empirical evidence, not on the basis of the number of papers.

Should national groups and ethnic groups be judged by this method when it comes to scientific credibility in any subject ? Should the indicators used in this study be normalized to the size of the population of the group in that case ? For example, can the group " Indian climate scientists " be compared with the group " US climate scientists " using this method ? After normalizing for population ? If the method leads to the conclusion that one of these national or ethnic groups isn't " cream of crop " in a particular subject, does this mean that the publications of any scientist belonging to that nationality or ethnic group should be ignored ? This question can't be brushed aside that easily since the method uses number of papers and citations as a criterion. How do you know which kind of groups can be compared using this method ? One could argue that only groups on two sides of a scientific controversy should be compared using this method ? But why ? Why can't it be applied to compare any kinds of groups ? And to produce rankings of groups ?

And if I remember correctly, the non-skeptics themselves underestimated things like Antarctic ice melting. So, the " skeptics " were bigger underestimators than the " non-skeptics ". At the end of the day, looks like both groups underestimated the importance of something that affects human future so critically. So, how can the " skeptics " be singled out for criticism ? This is a ridiculous and utterly wrong method. It can't be used to establish scientific credibility.