Thus reads the title of an editorial in Science Magazine by Cornell physicist Kurt Gottfried and former head of the NIH Harold Varmus in the March 20 issue. On behalf of the larger scientific community that is breathing a collective sigh of relief, the authors want to acknowledge the new regime in Washington. Said President Obama: “promoting science isn’t just about providing resources—it is also about protecting free and open inquiry…free from manipulation or coercion, and listening to what [scientists] tell us, even when it is inconvenient—especially when it is inconvenient.”

In the President’s Memorandum of Scientific Integrity he called for new standards “designed to guarantee scientific integrity throughout the executive branch” and to ensure “that scientific data are never distorted or concealed to serve a political agenda.” This is as it should be. It takes care of the “in-house” problem that we had within the Executive branch during the previous Administration, where manipulation and coercion were rampant.

Site map

The statement, however, makes the charming assumption that the studies served up by “science” are themselves value-free, which of course is not the case. It assumes that scientific reports are not biased, and that they don’t conceal a hidden agenda. This proposition has in recent years become very questionable indeed, with the increasing dominance of academic research by corporate funders that have very clear interests at stake. Additionally, we have the problem that individual scientists may also be personally either corrupt or simply too zealous to be appropriately self-critical in their work.

The problem of plain, naked scientific fraud is very difficult for the “system” to deal with. An analogy to the abuse issue among the priesthood may be helpful here. In retrospect one wonders why that issue was not taken in hand sooner by the Church hierarchy. The answer lies, I think, in the fact that the behavior of the deviant priests was so far outside of the paradigm that it could not be grasped. An appeal to the will of the offending priest to alter his behavior and to honor his covenants was clearly unavailing, as was the resort to “therapy.” That all took a long time to be recognized and acknowledged. If the hierarchy truly believes its message of the moral governance of one’s life, of the transformative power of faith, and of forgiveness and redemption, which it surely does, then one can begin to understand its blindness to what was actually going on.

Similarly, bald scientific fraud lies so far outside of the mind-set of the scientist that the research enterprise does not have good institutional means of dealing with it. Scientific fraud just doesn’t happen. And if it does happen after all, it’s a matter of the isolated rogue scientist. We don’t have a “rotten apple” phenomenon here that threatens to contaminate the whole barrel. No real institutional threat exists. In the scientific realm we operate with a belief in the self-correcting nature of the scientific process, and we assume shared fealty to the disciplines of the scientific method, a faith that may be just as touchingly naïve as that of the Church hierarchy.

Most likely, scientific fraud occurs on a continuum. We have slight indiscretions with regard to scientific rigor and probity at one end all the way out to wholesale data fabrication at the other. Incidence is probably broadly distributed as well. The problem is that the fraud—large or small—is very likely to be biased toward fulfilling the scientist’s expectations, and these expectations are likely to be aligned with the prevailing models. So the scientific process is just being helped along a little bit. If the result is to confirm what is already being expected and hoped for by the scientific community at large, then suspicions may not be aroused, and critical faculties may even be attenuated. On top of all the natural tendencies toward a confirmation bias an additional hidden variable may be at work. Scientific fraud is yet one more agency on behalf of the status quo in science.

Just as the Church for many years treated each case of sexual abuse as an isolated issue, we tend to treat each instance of scientific fraud as an isolated case. The possibility of a much larger, systemic flaw, was not considered by the Church for many years, and it is not being considered by “Big Science” now. And yet that’s probably what we have to contend with. With regard to issues such as vaccine involvement with autism, the fraud has become institutionalized. The institutional opposition to neurofeedback is another case in point. In those rare instances were fraud is apparent, the perpetrator should be held up to public opprobrium.

The Case Against Mark Nuwer
Mark Nuwer is a neurologist at UCLA who has been agitating against the reliance on standard QEEG-type of analysis in the discrimination of traumatic brain injury. His campaign has been ongoing for nearly twenty years. Most recently, he has been accused of misreprenting data to make his point, and from the court filings it appears that the evidence against him is pretty solid.

Instructions were given to Jack Johnstone of Q-Metrx, an innocent party in these proceedings, to analyze QEEG data in four different ways, only one of which was the conventional approach based on artifact removal prior to standard statistical probability mapping. In two of the runs, artifacts were to be left intact in the record. The results of those four different analyses on the same data set were then presented by Nuwer in court as if all were “legitimate” analyses that simply differed one from another as they might, for example, with different services analyzing the same data. The court was not informed that the different data sets were obtained under variable assumptions, most of which were contrary to standard practice. Three of the four traces are shown in Figure 1. (The fourth trace, being artifact-dominated, was identical to the first, and is not shown.)

The intrusion of fraudulent practices into this issue is doubly unfortunate because one’s tendency is to throw the hypothesis out with the fraud. If Nuwer’s case against statistical probability mapping (SPM) is fraudulent, then SPM must really be ok. There are, however, legitimate issues to be resolved with regard to discriminant analysis of the EEG for purposes of differential diagnosis. By making the case more of an absolute all-or-nothing proposition, Nuwer undermines the more subtle analysis of issues that must continue. Nuwer’s case against the use of QEEG in discriminant analysis has become an absurd caricature of the scientific process. It is hard to see how this will be cleaned up.

Science and the Courts
The legal history of the issue of discriminant analysis illustrates a larger point, which is that the courts are not a good way of resolving scientific disagreements. The tendency for people on all sides of a legal issue is for each to shore up the respective position to put the best light on it. The easiest means are by the selective presentation of data. The falsehood typically does not lie in the data themselves but in the withholding of other data that the court will know nothing about. If it is a challenge for scientists to be scrupulously correct with regard to data that support their fondest ideas, we cannot expect such probity to prevail in an adversarial setting.

The reason the issue of discriminant analysis spends so much time in court is traceable to insurance companies’ ongoing interest in fuzzing up the issue of traumatic brain injury. The big money, therefore, sits only on one side of the table. This is the second primary reason not to see the courts as a proper venue for scientific issues. Large financial interests can drag things out forever by the mere artifice of continually raising doubt about issues that are complex, and fraught with intrinsic uncertainties. This happened with the case against tobacco, and it happened with environmental issues generally.

Recently the proposal was made by the Chamber of Commerce that a Scopes-type trial be mounted on the issue of global warming. What a circus that would be. In the actual Scopes trial, let us remember, the issue to be resolved was not the validity of the theory of evolution but rather the question of whether Scopes violated Tennessee law in teaching the subject. It turns out that he had, and he was duly found guilty. The larger scientific issue of evolution is not something that a Tennessee court, or any other court, should ever agree to take on. The same goes for global warming.

The suggestion of a trial is surely badly motivated. Such a trial would continue endlessly, which is just what Chamber members would want. It would postpone the day that they might have to take action against global warming. We are seeing here yet another powerful institutional bias in favor of the status quo.

We retain the fond hope that corporations are the source of innovation, and so we allow them wide latitude in our society. The largest enterprises, however, find it easier to deploy their innovative skills in protecting their turf. When it comes to true innovation, corporations usually just acquire it after the hard work is done on a small scale by the real entrepreneurs. Corporate innovation is called for to separate the inventor from his invention. The large corporation is more like an aircraft carrier that cannot be expected to redirect quickly. Its priority must be to protect itself by means of a large group of support vessels. In the real world, one of the support vessels is our court system. Long ago, the Supreme Court gave companies standing in the courts with all the rights of individuals. They have become the first among equals.

Resources:
Information on the ethical complaint can be found at the following link:
http://www.appliedneuroscience.com/NuwerEthicsComplaint.pdf

Siegfried Othmer, Ph.D.
www.drothmer.com

This entry was posted on Wednesday, September 2nd, 2009 at 12:12 am and is filed under Research, Scientific. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.