Design Science is a problem-solving approach and actually has its origins in engineering. The approach is that you solve a problem by developing a prototype. Action research is the solution to a problem through observation, action and evaluation. Action research is another such approach and also aims to solve problems. This is done through an analysis, action, evaluation and adaptation in different iterations. I have recently seen action research especially for testing theories in practice, but this can also be used for prototypes. Supervisors often reject the design science methodology because it is still controversial and US-American. It is then advisable to use action research.

Both methods represent a counter to classical research. They make concrete interventions in the environment and use new methods to solve problems. The actual research is not the data collection but the way to the solution or the development of the prototype. At the same time, this is also the big criticism: You produce a solution that is quite subjective and strongly tailored to a problem.

In business informatics it is usually used to develop a software prototype. If you are programming software, then this approach can be useful.

Approach of the methods

Design Science consists of two processes: “Build” and “Evaluate”. The background is the creation of a prototype or software. Before performing Design Science, you should plan a system concept and the test environment.

Example of design science

In this example in the figure, we want to develop software for measuring service innovations, which was my bachelor thesis. What is important is not the prototype but the way to get there. So the search process for the solution. Incidentally, the design science can include a small literature analysis in advance, if necessary. The figure shows that the procedure often contains different iterations. In bachelor theses, however, I often experience that it is just an iteration.

To evaluate the software, you should combine another procedure such as an interview or questionnaire. In this way, the solution competence of the prototype can be ensured and feedback can be collected. You usually proceed in iterations: Prototype -> Feedback -> Prototype -> Feedback etc.


Clearly, both methods are very suitable if you want to develop a prototype or software for your thesis. Criticism lies in the “hands-on” mentality of the approach as a poorly defined scientist and a more practical approach. Some researchers therefore reject the design science methodology. I then recommend the more established action research to be preferred.

Both methods aim to solve problems through analysis, action, evaluation and adaptation in different iterations. An alternative could also be used in action research in testing theories in practice through a prototype.

For experienced programmers, the creation of software is a good way to get away from classic research methods in the thesis and to create something “to touch” but still use a scientific method.

[student] [fotolia]

I blog about the influence of digitalization on our working world. For this purpose, I provide content from science in a practical way and show helpful tips from my everyday professional life. I am an executive in an SME and I wrote my doctoral thesis at the University of Erlangen-Nuremberg at the Chair of IT Management.

By continuing to use the site, you agree to the use of cookies. more

The cookie settings on this website are set to "Allow Cookies" to provide the best browsing experience. If you use this website without changing the cookie settings or click "Accept", you agree to this.