Gladys Reynolds
Centers for Disease Control
This article is a follow-up to the fall Newsletter article by Dr. Marlene Egger on the American College of Epidemiology (ACE) Statement regarding minority health [1,2]. This article is excerpted from a talk I presented at a conference on "Statistics, Science and Public Policy" in April, 1996, in England, focusing on the importance of cultural diversity in counteracting cultural bias.
It is generally believed that scientists are bound by a set of values and traditions that embody honesty, integrity, objectivity and collegiality. Conventionally, scientists have relied on each other, on the self-correcting mechanisms intrinsic to the nature of science, and on the traditions of the community to safeguard the integrity of the research process. [3-5] According to this model, objectivity is embedded in the social structure of science.
However, in the political, economic and social context in which science is created and reviewed, the self-correcting mechanisms of science may not be sufficient. Unacknowledged bias is perhaps the most serious of threats to the proper functioning of the scientific method.
This has become especially clear to me while working in the area of minority health, an area particularly impacted by the social and cultural biases of scientists. As scientists, we must scrutinize our own social contexts and compensate for our own biases which are a serious threat to the proper functioning of the scientific method and which can occur in every stage of the scientific process. These stages include: 1) posing scientific questions, 2) conducting research and interpreting results, and 3) peer review, debate, discussion and communication of results.
The interests that underlie a particular piece of research and its results can affect the way scientific questions are posed, which questions get posed and which don't, whether there is healthy debate, and whether policy makers are receptive to the results. One important aspect of these interests is funding and sponsorship.
The research that is supported and funded is determined by the social, economic, and political contexts in which science is done. Knowledge is created in a social context and its creators are a part of society. There are many articles and books written on the history of human experimentation which indicate that the scientific community is as likely to embrace the biases and beliefs of the times as others and that it has allowed dogmatists and racists to use their status to further their personal agenda.[3, 6-15] Examples of this include the broad acceptance accorded the work of Samuel Morton, Paul Boca, H.H. Goddard, Lewis W. Terman, Robert Yerkes and Cyril Burt on racial, sexual, and cultural differences in human intelligence.[3,6,9,12]
There is much written about factors that are damaging to the integrity of the scientific process: errors in judgment, inappropriate design or methods of measurement or analysis, etc. [11, 16-19] There is much less discussion about the subjectivity of our social and cultural observations, about how our work is dependent on conscious or unconscious agendas--on values. Our observations are framed in a context of cultural and historical meanings and forces. [20,21] We depend on the power of our own observations in order to draw conclusions, but the act of observation is clearly subjective. We use our personal--yet culturally-constructed--experiences to both make our observations and to interpret what we have observed.[21] So it is not that we are influenced by political dogma, but that we are socialized into an acceptance of "common-sense" notions that are present in our culture. Thus, we bring issues of blame, of "innocent and not innocent" or "worthy and not worthy" victims, to our interpretations of sickness and health, wealth and poverty.[21] These cultural assumptions act as invisible-- and thus untested-- premises and as such conflict with scientific method.
Peer review, debate and criticism are crucial to the quality of science and the policies that are based on it. The desire to reproduce and verify results is largely motivated by the need to resolve debates. In the process of reviewing research, scientists again bring their own cultural and ideological biases into their evaluation of the adequacy of their colleagues' hypotheses, the collection of data, and the analyses and interpretation of results. Journalists and policy-makers often exacerbate these biases when they bring scientific information to the public.
The unique insight and experience of members of different minority groups are again needed to evaluate the appropriateness and importance of different studies and of prevention and intervention strategies, to interpret and disseminate results, both to their own diverse communities and in the scientific literature and to support and promote equal access to education, economic security, political and scientific input, quality health care, etc., for all members of our society.
One way to increase the rigor of our scientific process is to bring diverse viewpoints into the scientific community. These diverse viewpoints will provide checks and balances on the biases of individuals by bringing a diversity of viewpoints and biases to the discussion process. These cultural checks don't exist when scientists all share common values and backgrounds, as has historically been the case.
As I already noted, the attempts to measure intelligence earlier in this century provide an excellent example of the kind of bias that can go unchecked in all stages of the scientific process in the absence of diverse viewpoints. Scientists long assumed that races could be ranked on an evolutionary ladder, with whites at the top, and that this was reflected in differences in the physical attributes of humans. Leading scientists of the day, in their "measurements" of intelligence, confirmed the theory, initially, with measurements of the skull, and later with IQ tests. Though universally seen as without scientific merit from today's perspective, early intelligence tests which were used to "prove" that Jews, Hungarians, Italians, Russians and blacks were feeble-minded and were intended to help identify and curtail the reproduction of bad genes.[6, 7, 9, 13] The Eugenics movement, for example, was initially a recognized and respected branch of genetics that eventually led to the practice of mass murder. [6, 8, 14, 15] These tests had great scientific prestige among informed psychological opinion. Psychometric testing has had continued support throughout this century, despite many demonstrations that the premises of such tests have no scientific validity. [3, 6, 7, 12, 14, 22] However, IQ tests, SATs and similar tests are still heavily relied on to determine (or rather, justify) who does and who does not deserve access to higher education.
Other common-sense notions that come from cultural ideologies, affect our scientific investigations. The virtues of liberal individualism (competition, personal responsibility, self-reliance) and of capitalism (profit, privatization, self-interest, political influence) appear in our scientific research as much as they do in our public policies. Examples of this are studies that emphasize the behaviors of individuals as causes for poverty, criminal activity, low educational achievement, morbidity and mortality rather than looking at broader structural or economic causes. Such studies encourage the notion that such social problems are caused by the moral failings of individuals and resist an analysis of the social contexts in which they occur. Hence, we see a number of genetic explanations offered for complex social phenomena. Specific genes are sought for intelligence, for obesity, for homosexuality, for violence, and for mental illness.
As we work to understand and close the severe disparities in health outcomes experienced by people of color compared to white populations, our studies/interventions/recommendations need to be grounded in the context of the conditions and circumstances that the minority person deals with. Such health problems have a historical, political, economic, legal and social context. [22-26] The study of minority health crosses many disciplines (economics, anthropology, epidemiology, medicine, sociology, engineering, biology, criminology, psychology, statistics) and many areas of public policy (education, environment, public health, criminal justice, assistance programs, rehabilitation) which persistently embrace current prejudices in all aspects of policy-making. Unless we can frame the solutions in this context we have little chance of closing the gaps, such as the increased gap in infant mortality and age-adjusted mortality which have followed increases in income disparities.
Diversity within the scientific community can make the process of science less susceptible to errors that occur because of scientists' own failures to give equitable consideration to interpretations and conclusions that do not fit their own biases, by enriching scientific debates, and by giving politicians and other lay people confidence that their political views are well-represented in the scientific community.
As scientists, we must insist on asking what the motivating assumptions are that underlie the basic premises of the research? Are they grounded in preconceived or prejudiced notions (e.g., racism, classism, sexism, religious intolerance, homophobia, xenophobia)? And how do our cultural assumptions bias the science we conduct?
Then we must be willing to deal with these legacies, i.e. their cultural and historical origins of our cultural values and how these values have affected and will continue to influence every aspect of life and public policy, whether it is in science, education, the environment, criminal and civic justice, or health care access.
The unique insight and experience of members of different minority groups are needed at all steps in the scientific process: defining the scientific questions, conducting research, to evaluate the appropriateness and importance of different studies and of prevention and intervention strategies, to interpret and disseminate results, both in the diverse communities and in the scientific literature and to support and promote equal access to quality health care for all members of our society.
Because substantially different viewpoints presented by diverse groups contribute to the rigorous development of our thought and our integration of new and more inclusive perspectives, our professional organizations must make active efforts to improve the inclusivity in our professions. Our professional organizations should adopt statements of principles and goals that recognize the importance of minority input and the need for racial, ethnic, and cultural diversity in all the scientific disciplines because it is an ACTIVE corrective for past bias and puts on the alert to control for bias now and in the future.[1, 2, 21, 22, 26-29]