Test Driven Modeling: The Impact on Maintenance

The experiment was conducted in December 2010 at the University of Innsbruck in the course of a weekly lecture on business processes and work ows; all in all 12 students participated. To prepare the students, a lecture on declarative process models was held two weeks before the experiment. In addition, students had to work on several modeling assignments using declarative processes before the experiment took place. One week before the experiment, the concept of testcases and their usage was demonstrated. Immediately before the experiment, a short lecture revisiting the most important concepts of TDM and the experiment setup was held. The rest of the experiment was guided by CEP’s experimental work ow engine, leading students through an initial questionnaire, two modeling tasks (one with the support of test cases and one without the support of testcases), a concluding questionnaire and a feedback questionnaire. The experiment was concluded with a discussion to exchange students’ experiences and to revisit the experiment’s key aspects. A detailed description of the experiment can be found here.

Material

The complete assignments can be found here.

The assignment sheet includes a download link for TDMS. If you do not want to download the software, but still want to take a look at the provided process models, you may want to download the models here.

Finally, our students received a handout pointing out potential pitfalls and briefly explaining the constraints (to ensure that their semantics is understood properly). This document can be downloaded here.

Publications

  • S. Zugal, J. Pinggera and B. Weber: The Impact of Testcases on the Maintainability of Declarative Process Models. In: Proc. BPMDS ’11, pp. 163–177, 2011.