Fraud from the frontlines: the importance of being nice
- First Online:
- Cite this article as:
- Douglas, H. Metascience (2011) 20: 553. doi:10.1007/s11016-010-9492-2
- 50 Views
In this charming and engaging book on scientific fraud, David Goodstein draws on his extensive experience as a practicing scientist and vice provost at Caltech. The subtitle is appropriate—the book is full of Goodstein’s personal stories about differentiating scientific misconduct from acceptable science. Such stories, supplemented with an examination of Robert Millikan’s oil drop work, provide an incisive look at actual scientific practice.
Goodstein defines fraud narrowly (and correctly, in this reviewer’s mind) as fabrication of data, falsification of results, and plagiarism. Thus, the book is not a sufficient treatment of all the ways in which scientists can misbehave (there is no discussion of other unethical research practices such as mistreatment of subjects, problems raised by conflicts of interest, or the mishandling of sensitive results, for example)—nor is it meant to be. Indeed, the detailed examples of fraud are largely cases of data fabrication, and thus plagiarism is not addressed in detail. But it is a fun and quick read, provides a refreshingly honest and grounded look at fraud, and could be useful in a course for budding scientists on research ethics.
Goodstein opens the book with a candid look at scientific practice. He examines fifteen reasonable sounding maxims (which most philosophers of science will recognize as flawed at the outset if taken strictly) and proceeds to point out why scientists should not always adhere to them (6–15). Goodstein makes clear that although such maxims fail, the rules against fraud (narrowly defined) are strict. He then provides a really useful account of possible career paths in science, an account every graduate student should read as they think about what kind of job in academia they want (18–27). According to Goodstein, the drive for credit is the central motivating force for the scientist, as s/he scales the twin systems of Reward and Authority.
This is background for the main story, fraud in science. Goodstein draws some interesting conclusions about the nature of scientific fraud. He suggests that fraud is committed by scientists who “almost always believe that they are injecting a truth into the scientific record··· but without going through all the trouble that the real scientific method demands” (2). Thus, the attempt to deceive is usually not one of trying to get scientists on the wrong track, but rather of trying to take credit for work not actually done, but presumably that would turn out this way anyway. Scientists who commit fraud are also usually “under career pressure” and are working in fields where it is difficult to precisely reproduce results (such as in the biomedical sciences) (3–4). The examples of fraud in the book largely hold to these maxims, although the ease of reproducibility is the source of fraud in one case—the supposed discovery of element 118 by Victor Ninov (102–106). Because there was a theoretical prediction for how to produce the element and how it would behave, there was a general expectation that experiments would produce a certain result. Ninov claimed to have those results (hoping to get credit for the first production of element 118), but when other researchers failed to find the element, his fraud was exposed. It is an interesting twist on the usual story.
In addition to cases of fraud in biology (chapter 3) and physics (chapter 6), Goodstein discusses normal scientific practice (from his own area in physics) for contrast (chapter 7). He does not spend time on experimental details as he does in the earlier chapters, instead providing a clear introduction to high-temperature superconductivity in physics. The point made is that despite physicists not having a theoretical understanding of how or why high-temperature super-conductivity occurs, physicists can still readily agree that it does. He conveys the sense that nature can still surprise the scientist, indeed that that is part of the fun.
One of the most surprising chapters is the chapter on cold fusion (chapter 5). Based on his knowledge of the issue from a close friend who conducted research on the phenomenon, Goodstein uses the example to demonstrate abnormal, but not fraudulent, science. The difficulties of reproducibility come to the fore, as does the drive for the credit of discovery, and Goodstein leaves open whether there is anything to cold fusion.
The most controversial chapter might be the chapter on Millikan (chapter 2). As the acknowledged “all-around patron saint” of Caltech (29), Millikan receives both careful treatment and perhaps an overly quick exoneration. Goodstein’s discussion of the claims that Millikan mishandled his data is thorough and well worth the read, but Goodstein is focused primarily on Millikan’s oil drop experiments. Millikan’s dispute with Ehrenhaft, however, stemmed from his earlier water drop experiments. Because these were mere precursors to his Nobel Prize winning work (the move from water drops to oil drops was crucial because of the longevity of the oil droplets compared to water droplets), Goodstein discusses only Millikan’s most famous work and finds he did not misrepresent his data. Having exonerated Millikan of fraud, he gives him a pass on his other misdemeanors of mistreating grad students, being sexist and an anti-Semite (29–33 and 46, 47). Goodstein’s treatment of these issues is not central to his book (which is about fraud and non-fraud, not how to be a good scientist all around), but still, his eagerness to see no wrong from Millikan is disappointing.
Consider the plight of Harvey Fletcher, Millikan’s hapless graduate student, who probably should have shared the Nobel Prize with him. There are two published versions of how the researchers (then at the University of Chicago) moved from water to oil in their efforts to measure the charge of the electron. The first is from Millikan’s autobiography, where he claims that he thought of the change on the way back from a conference in Winnipeg in August 1909 (Millikan 1950, 75) and had a mechanic begin work on the apparatus at once. But this version is most likely pure fabrication because in October 1909, Millikan makes no mention of oil drops in a paper read at a conference (the abstract of which is published in December 1909 in Physical Review and the full version in February 1910 in Philosophical Magazine)—he is still using water drops—and he surely would have had some data by October given how superior oil was to water for the experimenters’ purposes.
Fletcher has the other published version—that he thought of using oil instead of water in December, and he built the first oil drop apparatus that month. Full results using oil drops were presented at the April 1910 Physical Society Meeting. It is these data that led to Millikan’s Nobel Prize winning paper and that Goodstein analyzes. Fletcher recalls doing much of the writing for this paper, but Millikan maneuvered Fletcher into granting Millikan single authorship of that paper in the smarmiest of ways—preying on the vulnerabilities of a young graduate student needing his Ph.D. and employment to support his new family (Fletcher 1982 lays out the story with graciousness toward Millikan). Millikan was a great experimentalist but he was not a great human being, and Goodstein could have been more unflinching in his assessment of his own institution’s patron saint. Indeed, Goodstein could have used Millikan to show that being a nice human being is neither necessary nor sufficient for being a great scientist.
The main challenge of the book underlies this treatment of Millikan—Goodstein is relaying stories from the “frontlines” of science. As I note above, most of the book is based on Goodstein’s personal experiences, cases that arose in Caltech, in Goodstein’s own research, or in the research of close friends. Indeed, the chapter that discusses the development of federal rules on fraud reads like a frontline soldier grumbling about the higher ups (chapter 4). Goodstein writes the story so that Caltech and its/his efforts look eminently reasonable and sensible, while the Office of Research Integrity (in its various incarnations) looks bumbling and foolish, to the point that I wondered whether the view from the frontlines was too close.
Indeed, it may be hard to see from the frontlines just how prevalent fraud is. Goodstein makes no definitive claims about the rate of fraud, but presumes that the “great majority” of scientists are honorable and never come close to fraud (xiii). A recent confidential survey of 2600 scientists, however, produced over 1000 reported incidents of fabrication, falsification, and plagiarism (Koocher and Spiegel 2010). Happily, researchers also found that over 2/3 of scientists reporting incidents attempted to intervene to prevent misconduct, and, often by taking a non-confrontational approach, were able to steer their colleagues away from misconduct. Fraud is not rare in science, but it can be kept in check by colleagues. Given how important collegial interactions appear to be in policing fraud, perhaps being a nice person is crucial to the modern scientific community after all.