Meaning and Measurement

 

Conceptual Basis of the Taxonomy

Qualitative/Personal

Evidence is primarily narrative in nature, emerges from a historical context, and is concerned with meaning and particular individuals. It involves the varying beliefs, attitudes, preferences, predispositions, perceptions and "epistemological frames" brought forth by both health care provider and patient.

Qualitative/General

Evidence is primarily social and historical, illustrates social views and preferences and is manifest in policy debates and consensus statements. Here, the managerial perspective emerges as does the set of sciences devoted to the study of health care in its organizational manifestations. The cultural, social and gender dimensions of evidence emerge more prominently in this quadrant.

Quantitative/General

Evidence is primarily statistical in nature and general in its application. Measurement and rigor are the chief concerns. Quantitative data derived through the application of recognized study designs constitute the basis of evidence. It is impersonal in the sense that the evidence is intended to be general in application and not subject to bias or self or group interest.

Quantitative/Personal

Evidence is primarily mathematical and personal. Research exemplars include some elements of clinical epidemiology, quality of life scales and cognitive psychology. Evidence is defined in terms of measurement of personal belief. The most clearly stated and developed manifestation of this is found in the development of Bayesian methods of reasoning.

Click here for examples of the evidence taxonomy quadrants

Back to Top

Epistemology

Locating Evidence

Scientific evidence related to health care can be situated or exemplified in relation to concepts of epistemology, the philosophy of science and logic. There is a good justification for using these frameworks to help us understand evidence.

What is Epistemology?

Epistemology is the branch of philosophy devoted to researching questions about the nature, scope and justification of human knowledge. More specifically epistemology asks questions about the nature of human beliefs, scientific and otherwise, and explores issues in the adequacy of reasons put forth to justify those beliefs.

     We have seen in the definition and characterization of evidence that there is a clear relationship between evidence and belief and the justification of belief. Thus there is a conceptual link between evidence and epistemology. Other disciplines, most notably sociology of science, also look at these relationships. However, in philosophy the varieties of possible ways of construing knowledge have been well worked out and include sociological, legal and scientific perspectives.

Knowledge and Evidence

What is the relationship between knowledge and evidence?

Knowledge can be defined as the set of propositions about any subject regarded as true. Modern scientific knowledge is quite complex and sophisticated. There is a lot of knowledge that is discipline specific. Discipline specificity is exemplified by the employment of unique vocabulary, technology, and patterns of practice. In health care these propositions would take a variety of different forms depending on the context in which the claim for knowledge was being made.

     It is unclear whether evidence is distinct from or synonymous with knowledge in the EBM debate. The terms are often used interchangeably, though in philosophy they are kept distinct. It is likely important to keep the concepts apart, as will be seen when the relationship of evidence to reasoning is considered in greater detail.

     The following example will help to illustrate the distinction. Imagine a physician and patient discussing whether or not a certain medication will be effective in lowering blood pressure. The physician states that she knows that hydrochlorothiazide will lower the patients blood pressure. The patient responds " How do you know?". Consider the range of possible answers available to the physician:

  1. Because I'm a doctor (Intuition/Authority)
  2. Because in my experience it works (Experience)
  3. Because it interferes with a specific biochemical process that will lower blood pressure (Basic Science)
  4. Because many well designed studies have shown that the drug is effective in lowering blood pressure (Clinical Science)
     In each of the cases above, a very different reason is given for answering the question how do you know. This illustrates the point that evidence is offered in support of knowledge claims, that is, evidence arises in the context of justifying claims and is distinct from the claims themselves. The practice of critical appraisal of data and the critical analysis of arguments is directed at assessing the adequacy of the justification that the evidence provides.

Back to Top

Aims of Science

Science has both theoretical and practical aims. Scientific evidence in health care operates at an intersection of both the theoretical and practical aims of science. The theoretical and practical aims of science can be broadly conceived as:

  • Establishing true propositions about the natural world
  • Providing a coherent account of natural processes
  • Providing explanations of phenomena that link observations to theories
  • Providing the justification for actions
  • Allowing for accurate predictions of future states of affairs

Scientific evidence in health care is largely practical in intent. It is intended to improve decisions and therefore outcomes in the health status of those upon whom evidence based medicine is practiced.
Lacey A. R., The Oxford Companion to Philosophy, Oxford University Press, 1995 by permission of Oxford University Press.

Back to Top

Methodology

Methodology is the manner by which observations are recorded and organized. In science, method is of crucial importance, it is meant to be explicit, systematic, and logical. From the adjacent diagram, it is clear that there is a great deal of complexity involved in scientific methodology. Method, theory, inference are intimately interwoven.

     The diagram helps us to locate the role of methodology in the creation of evidence in health care. In health care a wide range of designs are employed to generate evidence. Broadly speaking, they fall into two general categories: observational and experimental.

     In experimental methodologies, variables are under the control of the investigator and are manipulated through the employment of randomization. Examples of experimental designs are randomized controlled trials, N-of-1 trials, and many basic science studies.

     Observational studies can be quantitative or qualitative. In general, key variables are not under the control of the investigator. Quantitative observational study designs include cohort studies, case control studies and cross sectional surveys. Qualitative studies include interviews, document analysis, and focus groups.

     Each of these methods is designed to answer specific questions. Each has its inherent strengths and limitations. Understanding the relationship between evidence and methodology is crucial in order to properly interpret the results of a study.

     For an excellent step by step account of a quantitative study please go to On-line Training in Clinical Research with CORDS.

     For an account of qualitative study designs, please go to Qualitative Research in Health Care.

Back to Top

Interplay of Mathematical Models

Mathematics, Statistics and Evidence

     Most health care evidence results from quantitative analysis. Statistical methods have become the predominant means by which evidence is expressed. It is important to note that expressing evidence in terms of statistics is a recent phenomenon. Medical statistics only became a discipline in its own right in the 20th Century. Probabilistic approaches to medical knowledge have a long tradition, and medicine in ancient times was regarded as a stochastic art.

     Statistics is one form of quantitative reasoning. The adjacent figure, taken from the work of David Salsburg, shows how abstract mathematics relates to three distinct approaches: logical, deterministic and probabilistic. We will not go into these methods in much depth, but the important point to note is that probabilistic or statistical approaches are simply one particular way to approach quantitative evidence. We are reminded here of the elaborate sets of equations used in classical mechanics (The Newtonian System) premised on a deterministic universe that uses calculus as its language.
Back to Top

Probabilistic Models Further Refined

  This figure, also from David Salsburg, shows us how probabilistic models can be refined. Probability is a familiar concept to most people. We commonly use probability estimates in every day life and are surrounded by probabilistic reasoning in the media. A good example from daily life is the probability of precipitation found in weather forecasts.

     However, there is not agreement among statistical theorists about what a probability is. To one school of thought, a probability is an objective property of events in the natural world. This is often referred to as the frequentist position. To the other major school of thought, a probability reflects a measure of belief. That is a probability is a subjective property.

     The taxonomy approaches the issue of the meaning of probability by recognizing both definitions. It has been shown in research studies that medical professionals tend to interpret probabilities subjectively.

Back to Top