Martin Klasen, Saša Sopka
In medical education, there are often several didactic methods for teaching and learning a specific medical skill. For educators, there are often pragmatic reasons to decide for one or another of them, such as costs, infrastructural requirements, time expenditure or qualification of the teacher. However, a central aspect to consider is the learning outcome: Does a new method achieve a similar learning success as an established standard method? To answer this question, we need an appropriate method to assess comparability of learning outcomes.
In this paper, we present two essential statistical concepts that can address the issue of comparability of learning outcomes: Equivalence and non‐inferiority testing. We explain the ideas behind these concepts and illustrate them with an example data set. To clarify several concepts, we use theoretical examples from one selected field: the teaching and assessment of Basic Life Support (BLS).
Equivalence and non‐inferiority tests can be powerful tools for comparing teaching and assessment methods. However, their correct application requires adequate knowledge about their strengths, pitfalls and application fields. The aim of this paper is to deliver this knowledge and to provide clinician researchers with a practical guidance to a successful application of these methods.