Review of Methods for Conditional-Probability Computations in Medical Disciplines

Abstract

• There is a definite need for representations of, and tools for, conditional probability that enhance understanding, simplify calculations, foster insight, and facilitate reasoning. Such representations and tools are desirable and useful in a wide variety of scientific disciplines, but their utility in medical contexts and applications are stressed herein, so as to address a clinical rather than a mathematical audience. We employ a plethora of time-tested pedagogical representations or tools of conditional probability including: (a) Visualization on Venn diagrams or Karnaugh maps, (b) Reformulation as natural frequencies, (c) Entity interrelations via Signal Flow Graphs, (d) Specification of certain problem formats such as the format of Trinomial Graphs, and (e) Use of a pair of complementary Karnaugh-map-like or eikosogram-like diagrams. Most of the new representations or tools have well known histories of pedagogical advantages, but are still to be tested in the specific realm of conditional probability. Further assessment of the novel representations or tools proposed herein is needed. Each of these is to be taught to a group of students, and a control group of students is to be instructed via the conventional representation. Detailed statistical analysis of the outcomes is warranted. Similar investigations were performed earlier, but they were on a limited scale. Therefore, the need arises for an explorative study that exhausts all proposed representations or tools.