Iffy Science and Public Policy

George Rebane

[This is the transcript of my regular KVMR commentary broadcast on 18 April 2018.]

Will Rogers, humorist and keen observer of American life, once said, “It ain’t whatcha don’t know that worries me, it’s whatcha know that ain’t so.”

A lot of government procured science, especially science on which politicians base public policy, is not what Americans have been led to believe.  The mention of ‘science’ as supporting this law or that regulation is too often enough to stifle further critical examination of why we should stop doing this or start doing that, especially if you have never done science or are not that good with numbers.

While science has given us the technologies that make modern life possible, and often a pleasure, it is far from being a foolproof pursuit.  As we have noted before, science walks on two legs – Occam’s razor and falsifiability.  Occam counsels that the simplest explanation or theory is almost always the correct one.  And falsifiability demands that every scientific proposition or theory must be falsifiable.  That means it must contain elements that can be tested, and if the test fails then the theory is either rejected or raises serious questions that must answered by more experiments and testing.  And experiments that appear to have turned out well must then be replicated and their results reproduced by other scientists before credibility is established and the advancement is acknowledged.

Now also consider that at least half of the results in published scientific papers will subsequently be shown to be wrong, and we begin to understand why doing good science is really a careful, sometimes tedious, and a grueling slog.  This being the case, there is plenty of room for mistakes where science is involved.  And in recent years as more and more science has been done, we have run into more and more errors and mistakes, especially where science meets the politics of public policy.

The most obvious symptom of this is that we are informed of wonderous and remarkable discoveries and breakthroughs, on which new laws and regulations are often based, and later it turns out that the science could not be reproduced – it was either misunderstood or was simply a mistake, or even worse.

Too many so-called scientific results that politicians and regulators demand are fraudulent, delivered by self-serving scientists that bend experiment and /or the resulting data to match what their ‘customer’ providing the money wants.  This kind of practice has become an epidemic as reported by Drs Peter Wood and David Randall of the National Association of Scholars, an organization that researches the practice and delivery of higher education in America.  Their most alarming findings are about reproducibility – many of the important and accepted “landmark” findings in areas such as behavioral psychology, education, and climate change simply cannot be replicated by other scientists.  Nevertheless, their hyped results have been embedded into laws and regulations which determine how we raise our children, educate the nation’s students, and attempt to control our nation’s economy.

In a recent WSJ essay they present a summary of their research (here).  And since we know that the relevance and immediacy of every social problem is only told by its numbers, here are a few relevant ones.  The journal Science reported that only 39% of 100 prominent psychology studies could be replicated.  Amgen attempted to reproduce 53 “landmark” studies in hematology and oncology, and could only replicate six of them.  A 2015 study estimated that American researchers spend $28B annually on irreproducible preclinical research.  And on and on.

For the layperson and voter, the advice here is don’t immediately believe every new scientific finding, and then embrace the proposed law or costly regulation that some special interest wants to immediately build on such a hot-off-the-press report.  Instead, dig into the science a little bit, and at least ask your elected representative who else has been able to replicate the claimed landmark results.  After all, the bottom line will inevitably impact your health and wallet.   

My name is Rebane, and I also expand on this and related themes on Rebane’s Ruminations where the transcript of this commentary is posted with relevant links, and where such issues are debated extensively.  However, my views are not necessarily shared by KVMR.  Thank you for listening.

Comments

[Bill Tozer]

For those (including moi) who do not have the wherewithal to get around the WSJ paywall, here is a short article that shows some of the data, probably taken from the same WSJ article. This is not to repeat Dr. Rebane’s post, but to reinforce it. A companion piece, if I may.

https://patriotpost.us/articles/55454-the-lack-of-integrity-in-science-and-what-to-do-about-it

[Gregory]

Perhaps the poster child of bad research is the blaming of autism on childhood vaccinations. A number of scares have come and gone but some people are still blaming thimerosal, a mercury compound used as a preservative that some reasoned was to blame.

There may finally have been a reasoned solving of the puzzle... children's Tylenol (including generics) whose rise in popularity parallels the rise in autism being diagnosed.
http://journals.sagepub.com/doi/pdf/10.1177/0300060517693423

No, I would not immediately go on a crusade against Tylenol because of this research.

[scenes]

http://dailytorch.com/2018/04/national-association-of-scholars-calls-for-the-end-of-secret-science-in-government-regulations/

https://www.epa.gov/newsreleases/daily-caller-scott-pruitt-will-end-epas-use-secret-science-justify-regulations