Customer Reviews

On Fact and Fraud: Cautionary Tales from the Front Lines of Science

5 star:		(5)
4 star:		(0)
3 star:		(1)
2 star:		(0)
1 star:		(0)

 

 

I presented a "Case Studies in Scientific Ethics" course at the University of Maryland this past January, and I wish that this book had been available to me then. I will certainly use it next time. There is nothing like it, for it is written by someone who was quite close to the scene of action when scientific ethics became a focus of national attention in the late 1980s and early 1990s, and was engaged to some extent in spectacular episodes like the Baltimore affair. Moreover, there seems to me to be a prevailing impression - at least, I shared it before I designed a university course on the subject - that scientific ethics is a dry, compliance-oriented subject. NOT AT ALL, as this book makes most clear based on personal experiences of the author. In fact, the subject involves the intersection of a still imperfectly-understood field of philosophy, with one of the most dynamic and competitive arenas of human endeavor, scientific research. There are many "obvious" prescriptions that can be stated about ethical conduct of science, which are shown here, surprisingly, to fail in practice or in respect of their larger consequences. Much fruit for thought, pointing to the indispensable value of critical thinking informed by human experience. I rank this as the best general science book that I have read in some years. Its focus is on scientific ethics, but even those who have no particular interest in that subject will get valuable insight into the culture and practice of modern science.

David Goodstein is a professor and former vice provost at Caltech, and led the development of Caltech's rules of conduct. In this volume he discusses the philosophy of honesty in research, the challenge of reducing it to enforceable rules, and presents some illustrative case studies.

Before addressing what this book is, let me address what it is not. It is not essentially a popularization of science, or even of the cases it discusses. While some casual readers may find it interesting, it is really addressed to active researchers and those who study ethics. Goodstein is a good writer, and a notable science educator, so his style is accessible; but the book is a serious one, and best suited to a reader concerned with the details of research ethics.

The author lays out the history and general concept of research ethics. He then sets up a strawman set of nice sounding rules, only to explain why they are deficient. He goes through cases of true fraud, of fraud accused but ultimately not committed, of bad science that was not fraud (cold fusion), and finally too-good-to-be-real science that was actually real (high temperature superconductivity).

Goodstein offers a deep exploration of the difficulty in defining fraud in the real--and complicated and not straightforward--world of research as scientists actually conduct it. Sometimes there is a fine line between filtering data and misrepresenting results, sometimes there is a fine line between emphasizing the result the researcher believes to have achieved and de-emphasizing contrary evidence. Goodstein addresses the real world with solid understanding and experience, and with practical advice.

In most human endeavours one can find situations in which there is evidence of cheating, if one looks hard enough. And, as this fascinating book makes abundantly clear, science is no exception. After explaining how science works, reviewing the teachings of some philosophers of science and identifying "three conditions that are generally present when scientific fraud occurs", the author discusses, through six chapters, several cases in which questions of scientific misconduct or scientific fraud had been raised. Through careful analysis of each case, he shows that some of them were definitely not instances of scientific fraud - in fact, some turned out to be unexpected legitimate discoveries. Particularly captivating for me were the chapters on cold fusion and high-temperature superconductivity.

The writing style is clear, friendly, authoritative, widely accessible and quite captivating. The author's own personal anecdotes, peppered throughout the book, add a valuable special touch. Reading this book would be a definite asset for any research scientist, no matter what their field of expertise or their seniority. However, the book can also be thoroughly enjoyed by anyone interested in gripping stories from the front lines of scientific research - illustrating how science works and what constitutes scientific fraud (and what does not).

This is really an excellent book. It's particularly timely given the growing distrust and misunderstanding of how scientists work, particularly in the climate science arena. I wish that the author had addressed this topic; it probably became an issue too late in publication. But what he does have is illuminating, and contrary to one of the other reviewers here, I think the book is easily accessible to anyone with an interest in research and a free afternoon.

The book starts by establishing some basic ground rules on how the scientific method is supposed to work, and how it actually works in practice, given that scientists are human and not entirely disinterested parties when it comes to their theories and discoveries. The author then follows with a few examples of what does and does not constitute fraud. I really wish there had been more case studies here, but as he says in the introduction, he's limited to reporting the details of things he knows about firsthand. Most research fraud is not widely publicized. He concludes the book with a rather dry (but still interesting to people in the field) copy of Caltech's policy on research fraud and its investigation.

The author (and some of the reviewers here) identify the book's target audience as ethicists and scientists, but I actually think this book deserves a wider audience. It's not difficult reading, and it brings a little objectivity to a discussion (the motives of scientists) that is only becoming more polarized and strident.

It's not exactly accessible to a lay-person which is fine. I wasn't expecting anyone to be able to explain so many highly advanced scientific concepts in one small book. Some of the first-hand stories are nice. And Mr. Goodstein's sense of humor comes through with with some anecdotes. My main problem is that it also seems to be packed with filler. The photographs of scientists through out remind me of a high school student trying to pad out a research paper. They weren't really necessary. What I was really missing was main theme. Which I'm assuming is fraud does happen, but it's complicated. I would have enjoyed a book like that very much. But this book seemed to jump all over the place with stories, science, actual fraud and a republished lecture. The final chapter rehashed everything that was already covered which also wasn't necessary since the book is only 135 pages after take out the appendixes.

David Goodstein is a physicist at the California Institute of Technology who was for a time head of its science police -- not what the school called it, but that's what it was. At the time of his appointment, he says, he thought physics was mostly immune from deliberate fraud. Scintillation counters don't lie.

Much more problematic, he thought, was biomedical research, where results were necessarily mushier and where it was harder to control for every variable.

He quickly found that physics isn't immune from deliberate fraud when two cases presented themselves at his own institute. That shock set off a long rumination, resulting in this excellent little book, "On Fact and Fraud," which challenges some of the conventional notions of where the line lies that separates good from bad or real from phony science.

He uses real examples, from cold fusion to Milliken's measurement of the charge of the electron to high temperature superconductivity. This lets him make the point that science is the investigation of the unknown and, as such, there can be sharp surprises.

The discoverers of high temperature superconductivity were such mavericks that they committed a kind of fraud -- they concealed what they were doing, because it seemed too cranky to conventional thinkers. Yet they avoided what Goodstein labels the only kind of scientific fraud, which is the manipulation of what happens in the laboratory.

That they recorded scrupulously, and -- surprise! -- they found what they were looking for and won a Nobel Prize. This must give everyone pause.

But the story of superconductivity is so unusual that for a first cut, plain old ordinary cheating and self-delusion will suffice as usual suspects in most other cases. A good deal of thought went into the problem of scientific fraud in the 1990s, and Goodstein presents a list of 15 ideas about it that gained wide currency. All unworkable, he says.

Goodstein defines scientific fraud narrowly: "Scientific fraud consists of an explicit and well-defined act: faking or fabricating data or plagiarism."

He does not mention climate science. I believe he misses a kind of constructive fraud we have come to see too often there: Withholding of data or methods, cherry-picking (sometimes valid, sometimes not valid) data to sell a result and claiming scientific validity for models, as opposed to observations.

This may not constitute the kind of fraud that a college's science cop can investigate in a legalistic manner, but it presents a question of integrity far more consequential for the people who consume science than the most sedulously faked lab notebook.

If the lab notebook was confected, and if the apparent results were important, that eventually will out. (And, as we now know, some of the leading climate modelers, like Stephen Schneider, claim not to even keep lab notebooks, which would seem to fit Goodstein's definition of fraud in reporting.) The other kind of science fakery may not be revealed, at least not for a long, damaging time.