So how does one control bias in research? Over these intellectual and non-pecuniary sources of bias, the scientific method itself may be the best form of control. Indeed the purpose of following the scientific method, with appropriate controls, often blinded, is just that -- to ensure that the hypothesis can be objectively studied.
Financial interests are tangible and therefore inherently more accessible than these other interests. It is also true that, at least in some instances, they have played a significant role in encouraging inappropriate conduct on the part of researchers. It is therefore important that trainees consider, with other sources of bias, the question of financial interests and means to control their biasing potential. Of all topics in the responsible conduct of research, this may be the most difficult for trainees to appreciate. Stock holdings and large honoraria may seem rather exotic considerations for the young person at the beginning of a career and still struggling to meet student loan payments. Yet these issues may some day become more personally relevant, and in the meantime, it is useful to consider them in the context of the research of others. Disclosure is often cited as the key to dealing with conflicts of interest, and disclosure works by bringing scrutiny to the work at hand. Trainees, therefore, should be sensitized to the need to apply an appropriate level of scrutiny.
Case G1 presents a clinical researcher with a financial interest in a company whose product she will evaluate. This is the first level of conflict. She is provided with an additional financial incentive to enroll patients, another potential conflict. Although we may be reassured that, because this is a double-blinded study, her objectivity will be maintained, through the questions the reader considers the possibility that the blind may be broken by some unanticipated side-effect of the drug. In Case G2, a researcher is doing evaluative research for a cosmetics company while knowing that the company is anxious not only for quick results, but for results that will validate standing claims concerning a product. An agent of the company has informally suggested that positive and timely results will lead to continued funding opportunities for the investigator. Thus, the reader must consider the appropriateness of this interaction and how it might compromise the quality of the investigator's work or promote diversion from his other responsibilities.