Using a matrix-analytical approach to synthesizing evidence solved incompatibility problem in the hierarchy of evidence

By: Walach H, Loef M
Publication Name: J Clin Epidemiol
Year: 2015
Citation: J Clin Epidemiol. 2015 Jul 4. pii: S0895-4356(15)00321-2. doi: 10.1016/j.jclinepi.2015.03.027
Type: Journal Article



The hierarchy of evidence presupposes linearity and additivity of effects, as well as commutativity of knowledge structures. It thereby implicitly assumes a classical theoretical model.


This is an argumentative article that uses theoretical analysis based on pertinent literature and known facts to examine the standard view of methodology.


We show that the assumptions of the hierarchical model are wrong. The knowledge structures gained by various types of studies are not sequentially indifferent, that is, do not commute. External validity and internal validity are at least partially incompatible concepts. Therefore, one needs a different theoretical structure, typical of quantum-type theories, to model this situation. The consequence of this situation is that the implicit assumptions of the hierarchical model are wrong, if generalized to the concept of evidence in total.


The problem can be solved by using a matrix-analytical approach to synthesizing evidence. Here, research methods that produce different types of evidence that complement each other are synthesized to yield the full knowledge. We show by an example how this might work. We conclude that the hierarchical model should be complemented by a broader reasoning in methodology.