Rethinking the Threats to Scientific Balance in Contexts of Litigation and Regulation

Article excerpt

BACKGROUND: Although existing literature does discuss difficulties of doing science in contexts of litigation and regulation, work to date reflects little first-hand experience in such contexts. This gap is understandable but also potentially troubling: Concerns that seem important from afar may or may not match those that are most salient for scientists actually engaged in such work.

OBJECTIVES: Drawing on experience on scientific committees and in lawsuits, and using skills developed through doing qualitative fieldwork, I reanalyze past experience and field notes from the perspective of the 2006 Coronado Conference "Truth and Advocacy in Contexts of Litigation and Regulation." Although I initially shared the kinds of concerns generally stressed by other scientists and science-studies scholars--emphasizing overt, relatively sinister efforts to limit scientific objectivity--neither the literature nor my initial instincts provided adequate preparation for more subtle influences, which actually created stronger pressures toward bias. Particularly unexpected pressures came from consistent deference and praise for independence and credibility.

DISCUSSION AND CONCLUSIONS: The cases discussed in this article are by nature suggestive, not definitive; additional research is clearly needed. Future research, however, needs to focus not just on pressured toward bias that are easy to imagine, but also on those that are easy to overlook.

KEY WORDS: awareness of pressure, ethnography, power relations, scientific biases. Environ Health Perspect 116: 142-147 (2008). doi: 10.1289/ehp.9988 available via http://dx.doi.org/ [Online 7 November 2007]

**********

Most scientists are well aware of the need to guard against potential sources of pressure toward bias, particularly when work is conducted in contexts of litigation and regulation, but it is not clear that the usual sources of concern are actually those that are most important. In this article, I argue that, although overt pressures to slant findings may well be problematic, more attention needs to be devoted to the insidious but potentially more significant pressures toward bias that go largely unnoticed, often because they come from unseen or unexpected directions.

I present the argument in four main sections: In the first I discuss my experience on a scientific review panel, illustrating that potential sources of bias in science are more complex than is often assumed. In the second section I discuss that experience and this article's larger points in the context of existing professional literature on the topic, noting that the literature offers valuable contributions but also includes important oversights and omissions. The third main section illustrates this point by drawing on another, more recent experience, in which I was able to observe first-hand one of the ways in which a major multinational corporation was actively seeding the scientific literature. Finally, the fourth and closing section offers an initial or draft typology of key ways in which the unseen sources of potential bias may be considerably more significant than those that are seen and/or actively resisted.

Biases--Seen and Unseen

The committees of the National Academy of Sciences/National Research Council (NAS/NRC) generally begin with closed-door sessions, in which committee members are asked to disclose and discuss their sources of bias or conflicts of interest, including any business, research, or other interests or positions that might be perceived by outside observers as creating a potential for a conflict of interest. As NAS staffers commonly explain, the most knowledgeable scientists available on many issues also happen to be the ones who have worked and published extensively on the topics in question, so the intention is not to exclude all scientists who might have strong viewpoints. Instead, the goal is to have a balance of viewpoints and experiences and to discuss openly any such potential sources of bias, real or perceived, at the outset. …