WHEN DOES EVIDENCE BECOME SCIENTIFIC KNOWLEDGE?
By PREM MISIR, Ph.D.
Guyana Chronicle
September 15, 2003
Today, it has become commonplace for research findings to be seen as evidence which is treated synonymously as scientific or objective knowledge.
And today, we still have researchers who view their findings from a singular study as "God's word," or written in stone. But these results are not really scientific knowledge, or even objective evidence explaining a social problem, at least not as yet. Why? Let's take a hard look at some fundamentals of the scientific process.
Thomas Kuhn's (1962) THE STRUCTURE OF SCIENTIFIC REVOLUTIONS, showed that the empirical (observed) data researchers collect make up a selective viewpoint, and is not an accurate representation of the true external reality of society. Kuhn believes that scientists have to work with fundamental and varying points of view (paradigms), in order to observe the reality of the society. These points of view tell us how the society functions, or how it should function. Therefore, these diverse lenses are needed to see the society as it is. We cannot and must not accept mere personal viewpoints brought to any study.
Personal views brought to the study Prior personal views of reality usually determine the type of data collected. We could have a view whereby the unity of a society is based on mutual respect and acceptance of each other's culture. This view, then, would influence data collection, aimed at showing the importance of the multicultural curriculum, affirmative action, and equal opportunity employment. However, these views may partially blind us to other means of understanding the total reality of societal unity. Any insular and exclusive view also would erroneously impact the interpretations and conclusions derived from the data collected on societal unity. In effect, different explanations can be given for the same data.
For example, examine the debate over whether human development is inherited or learned. Some studies like Murray and Hernnstein's THE BELL CURVE, suggest that human development is inherited through the genes.
Reactions to these findings were very different. Some scientists believe that human development depends upon genetic factors. Other scientists that supported learning as the causal factor downplayed the results. These separate interpretations of the same data show the competing and conflicting views and beliefs about the importance of genetic inheritance versus learning.
Creating an awareness of these differing points of view would induce scientists and others to seek out for themselves any bias hatched by one group of views over others. In effect, a scientific explanation is not selected solely on the basis of empirical evidence, that is, the data collected. Scientific explanations are derived through a community of interactive views or paradigms, extracted from several studies, and not from one view drawn from a mere single study. In other words, scientific or "objective" knowledge requires for its growth, much more than our personal views of society.
Consensus on differing views to be applied to any study The evaluation of research findings through other studies' results, involves a negotiated and a competing process. What comes out of this evaluation is not total scientific knowledge, but a negotiated objective knowledge.
What I am saying is that objective knowledge refers to agreed-upon views, demonstrated through several studies at a point in time, which can change at another point in time. Indeed, scientific explanation and understanding change over time.
Our sense of a situation, at any one time, is indicated by limited views which fail to provide knowledge of the total objective reality.
Creating a total objective knowledge depends on making open the views and values we use in a study. This openness in making our views known will alert other scientists against those paradigms causing a biased examination and interpretation of the evidence.
Over the years, we see several research reports coming out of the Caribbean and Guyana. The findings from any one study itself are not sufficient to qualify as objective or scientific knowledge, unless the study incorporates an integration or modification of the available points of view from other studies. We need to know what points of view were used, and these paradigms have to be made public when findings are released. Researchers naturally bring their views to the research process. However, these views have to be integrated with the views extracted from similar-type studies, aimed at reaching a consensus on such views, or instituting some modification to the integrated views.
Reducing bias through integrated views - the public opinion poll Example Designing a survey on public opinion in Guyana landscaped with ethnic insecurities and sensitivities, and some political shenanigans, would, hopefully, make it mandatory for researchers to reach this consensus on competing and even conflicting views on the subject matter in question.
For instance, in using a questionnaire or an interview in a public opinion study, we have to ensure that questions are placed in uncomplicated, unemotional language, and have to be phrased well, so that all respondents will understand the questions in the same way. Question ordering also is important to avoid poll bias. This consensus on competing views from other similar-type studies facilitates the framing and ordering of appropriate questions in the questionnaire.
Further, poll results have to be weighted. The weighted process is applied to consider unequal probabilities of selection to make right the sample demographics. We also should guard against unduly manipulation by weighing to produce desired findings. Sampling error can be another source of bias, but not the most significant source of error. We use sampling error because it is one of the few sources of error that can be quantified. Even when all these errors may be addressed, we still need to consider the laws of chance which indicate that the findings of one poll out of 20 may be skewed merely due to sampling error. Findings from any study are mere sample findings, unless we can demonstrate that there is external validity, meaning that the results can be applied to the population at large.
Incidentally, polls, generally, are not administered for the good of society. Polls are conducted to produce a good story or to promote a cause.
For this reason alone, it is important to know who funded the poll and why it was conducted. These answers would demonstrate the different points of view available to explain a particular reality and why it is necessary to reach a consensus on these views, purporting to explain and resolve some problems in the society.
Personal viewpoints are selective Indeed, therefore, a particular social problem will have competing explanations for its etiology as well as its solution. Findings devoid of this scientific exposure aimed at reaching a consensus on differing views, would have a short shelf life, therefore, not allowing for the formulation of any meaningful social policies.
People must know now that research-type findings without a community consensus of the points of view used in the study, is not scientific knowledge. No hasty conclusions on the findings, therefore, should be made.
The worst-case scenario for the people is to readily accept any evidence, coming from any quarters, without the appropriate scrutiny and consensus on the points of view applied in fashioning the research questions.
People must demand to know what personal points of view are used to guide any study, as definitive interpretations and conclusions are drawn from these research results. These personal points of view may paint a selective viewpoint which does not cover the total reality of the society.