Research Methodology and Biostatistics Series II - Quality of Medical Research

Medical research has mushroomed, and concerns are expressed regarding the quality of such voluminous research output. Among several factors, the quality can be substantially improved by effective control ofuncertainties that severely afflict the reliability and validity of all empirical research. Uncertainty is an endemic affliction of human activity, and medical research is particularly susceptible. Two primary types of medical uncertainties are aleatory and epistemic. The former is inherent due to biological, environmental, and other natural variations, and the latter arises mostly from knowledge gaps. Biostatistical methods are equipped to handle sampling fluctuations that are a source of a major part of the aleatory uncertainties. Tools such as scoring systems, aetiology diagrams, and expert systems can help to reduce some types of epistemic uncertainties. Proper choice of research toolsremains crucial for controlling most epistemics.