Bias in Science (from Multiple Sources)

Related articles

Is bias a problem in science and science reporting? Well, sure. The desire to vindicate one's presuppositions plagues all areas of human knowledge. What makes science special, though, is the lengths to which its practitioners go to avoid letting bias affect their results and the willingness of its practitioners to subject their claims to merciless and thorough critical analysis by others.

In science, there's no falling back on faith, appeals to emotion, conspiracy theories, the subjective claim that "this is just how it seems to me" (the defense of such fabulist memoir writers as Robert Reich and Rigoberta Menchu), the equally mush-minded appeal to mass opinion, or pleas for privacy as a means of concealing data. (Columnist Jonathan Rauch notes that this openness and willingness to have one's work double-checked by others is a hallmark, in analogous ways, of science, markets, democracy, and well-done journalism and I would add academic philosophy to that list.) Science, contrary to the impression given by its detractors, is the least arrogant form of human knowledge, always starting from the implicit premise: "I may be wrong, so I offer what evidence I have in hopes that others will verify my findings or show me where I went wrong."

But bias creeps in nonetheless, and it can't all be chalked up to one cause, such as corporate evil, environmental activism, or superstition. First, some examples of the way that the profit motive can skew science:

Scientists, activists, and pharmaceutical companies sometimes bypass peer review by taking their cases directly to the media or to industry-funded medical conventions, where they can make presentations about new drugs to audiences of doctors.

I've heard first-hand accounts from people who observed the faking of data in drug trials, so surely the phenomenon is not all that rare (or else I just happen to know people who move in shady circles).

Companies like ADM, a maker of soy products, have an incentive to buy into any organic foods/holistic medicine boosterism in favor of their products (and HealthFactsAndFears.com criticized them for it, which should be a reminder that we care more about science than sticking up for corporations, contrary to some of our detractors' claims). The ADM soy case is a reminder that there's no clear boundary between corporate and "alternative" junk science.

A January 2002 Journal of the American Medical Association report warned of the temptation for doctors to talk their patients into participating in clinical trials (for which the doctor gets a payment), which may create the misleading impression in patients' minds that the doctors believe that the new method being tested is the best treatment for the patients' conditions.

In the same issue, JAMA questioned the growing interdependence between university researchers and industry (but didn't worry too much about funding from government, a topic I'll return to in a moment).

It is not without reason that people ask the question, "Who funds you?" A report in the January 24, 2002 American Journal of Public Health suggests that scientists demonstrably do tend to arrive at different conclusions depending on who funds them: "studies have found a strong association between author's opinions and their financial affiliations." The desire to "follow the money" is a legacy of the Watergate investigation and post-Watergate journalism. Unfortunately, the desire to follow the money is all too often an excuse for not following the underlying argument.

This is particularly true in politics: I recall an article that traced the funding of conservative magazines back to conservative foundations, making no effort to argue with the positions taken by these groups but implying that the mere movement of money was somehow corrupting, though many political magazines including left-wing ones such as The Nation are funded by non-profits. Furthermore, money never strikes people as dirty when it's funding their own side. You're not likely to hear left-leaning, follow-the-money journalists expressing concern about the superwealthy Ford Foundation, which not only funds domestic political causes such as the NAACP, feminists, and immigrant groups but also China's State Family Planning Commission, which enforces that country's one-child policy (the Chinese population control authority got some $640,000 from the Ford Foundation in 2001 alone).

So, while following the money can sometimes be instructive, let us keep in mind the numerous other biasing factors that science must resist:

While corporate money is typically regarded as corrupting, few people worry about the long-standing influence of government and foundation money over researchers and universities. It is curious that the government is still regarded as the gold standard of objectivity, even after the recent public debate about whether the government should place false news stories as part of the propaganda war against terrorists a debate that cannot have done much to shore up our credibility with overseas critics and domestic conspiracy theorists. Governments and foundations have agendas, too.

The omnipresent (but largely unperceived) leftward tilt of the media results in partisan groups such as Public Citizen being called "consumer groups," without elaboration, as if their recommendations would unequivocally benefit consumers, while conservative groups are usually labeled as such, and ACSH's partial funding by industry is often noted. This makes left-wing and environmentalist groups appear more objective and mainstream than their opponents, giving them a disproportionate influence in public debates, which can in turn affect research funding and policymaking.

Needless to say, the press is also biased in favor of scary stories as opposed to reassuring ones. The AP recently used the headline "Study: Cancer Rates to Double By 2050" to announce a study that in fact said that cancer rates are declining but (since cancer is often associated with old age) that cases will increase as a side effect of the aging of the population. (At least Reuters got it right in the their headline: "Cancer Rate Falls but Numbers Set to Rise.")

Publications with titles such as Food Additives and Contaminants Journal reveal in their very titles the underlying philosophical biases of the publications' founders. Consumers Union recently published a report in that journal about pesticide residues on produce, showing that they are slightly greater on non-organic produce but is a journal about "additives and contaminants" likely to question the underlying assumption that additives and chemical residues are harmful and that food "untainted" by humans would be superior?

Finally, there is simple pride, the fanatical devotion to a given pet theory that keeps cranks the world over revising and revising their designs for perpetual motion machines or their equations showing that the motion of the stars can be explained without recourse to traditional notions of gravity.

An article in the January 24, 2002 New England Journal of Medicine suggested the adoption of new guidelines for medical researchers that would instruct them to fully disclose any conflicts of interest that cannot be eliminated. That's a fine idea, but in the end, the only real defense against bias will be a combination of researchers' integrity and more important the world full of eager critics who pore over any new study hoping to find its flaws. The only realistic defense against bias is the countervailing effect of other researchers who also bring biases to bear on their work but in the end make possible a fairly reliable scientific consensus. That's not a radical new proposal for getting at the truth. That's science.

line
Responses:

May 28, 2002

The details you relate to the reader are uncontestable. Here in the Programme in Cultural and Media Studies at NU (the University of Natal, South Africa), this is an issue that is especially relevant in the light of some internationally embarrassing non-debates about the links between HIV and AIDS.

However, your conclusion that the "only realistic defense against bias is the countervailing effect of other researchers who also bring biases to bear on their work but in the end make possible a fairly reliable scientific consensus," is simply to replace one set of commercial or social biases with another.

Let's look at this in a slightly different way. Where, today, is the practising scientist in eugenics or phrenology to be found? Basically, nowhere except as the personal aberration of slightly dotty individuals. Why? Simply because there is no basis in independent reality for any argument that supports these pseudo-sciences. Yes indeed, the role of peer review is important in this regard. However, I would agree with scholars of C.S. Peirce's work that in the long run it is an indefinite community of inquiry based on its members' engagement with the independent reality of their subject matter that makes some line of inquiry science and not merely a play for power, prizes, or riches.

Our present debates about bias in science often fail to distinguish between the communities of scientific inquiry, on the one hand, and of academia on the other. To be an academic is either (1) to be involved with all the committee and teaching work of tenure or (2) to be a non-tenured academic worker striving to gain the long-term security of tenure. These are not the same as the striving after new facets of the reality of (say) retroviral reverse-transcription or the physics of quasar plasma jets. Our gaining and retention of tenure belong to the reality enshrined in a different set of ends or goals. We communicate as academics for the accomplishment of more immediate goals. The communication of a true community of inquiry, on the other hand, may involve questions and responses that are generations apart.

We are carnal, incarnate beings who, to be eligible for membership in our societies, must both defend and moderate our interests, making them compatible with the continuation of the greater social and cultural aggregation that is humanity itself. This immediate end is dependent on consensus, but the scientific end is uncovered (revealed, made manifest, etc.) by what Peirce scholars Christian Kloesel and Nathan Houser call "the resistance of reality to error." We do indeed need peer review, but not as a form of consensus (which Peirce would have identified with the "a priori" method of fixation of belief); peer review is, rather, one aspect or facet of the long-term business of learning from experience.

Arnold Shepperson
Graduate Programme in Cultural and Media Studies University of Natal, South Africa