Science between evidence and illusion

The ever-increasing number of scientific publications lets the half-life of evidence appear shorter and shorter. In scientific research, evidence does not enjoy 100% certainty. Considering the variety of approaches, points of view and constructions with regard to the comprehensibility and perceptibility of evidence, systematic error in scientific studies is immanent. In general, evidence should be well-supported by repeated implementation studies rather than by a single falsified hypothesis in a stand-alone study. Ideally, the estimated probability of certainty should be specified regarding the accuracy of observations, measurements and calculations, and conclusions. A division into categories of certainty may be useful, ranging from obvious (observational studies) and proven (interventional studies) to evident (meta-analyses and systematic reviews). The term evidence as currently used in scientific reporting might provoke higher expectations regarding levels of certainty than are justified. Evidence and bias have to be seen in a complementary context.