Evaluations in the Science of the Artificial – Reconsidering the Build-Evaluate Pattern in Design Science Research

The central outcome of design science research (DSR) is prescriptive knowledge in the form of IT artifacts and recommendations. However, prescriptive knowledge is considered to have no truth value in itself. Given this assumption, the validity of DSR outcomes can only be assessed by means of descriptive knowledge to be obtained at the conclusion of a DSR process. This is reflected in the build-evaluate pattern of current DSR methodologies. Recognizing the emergent nature of IT artifacts this build-evaluate pattern, however, poses unfavorable implications regarding the achievement of rigor within a DSR project. While it is vital in DSR to prove the usefulness of an artifact a rigorous DSR process also requires justifying and validating the artifact design itself even before it has been put into use. This paper proposes three principles for evaluating DSR artifacts which not only address the evaluation of an artifact’s usefulness but also the evaluation of design decisions made to build an artifact. In particular, it is argued that by following these principles the prescriptive knowledge produced in DSR can be considered to have a truth-like value.