Abstract:

In the industry, software testing is considered a de-facto technique for assessing the quality of software.

However, lecturers of Computer Science programs struggle to include software testing in their courses.

This is due to a myriad of factors, one of which is the lack of a fully developed body of knowledge on the didactics of teaching software testing. In our previous work, we identified that most students employ a so-called `developer approach' when modelling test ideas for a software system. This approach shows that students primarily use conceptual knowledge gained from programming courses in their testing strategy. These students see testing as a form of problem solving and apply an approach stemmed in rationalism. We believe software testing should not be taught solely from rationalism but should include much more empiricism. Tests should be small scientific studies where students use heuristics and exploration to formulate hypotheses about the expected workings of the system, and then conduct an experiment to find answers to these hypotheses and evaluate how the system responds. To shift the student's mental model to empiricism, we developed a serious offline game, where players devise a testing strategy for a system under test through dynamic gameplay. The objective is to reach a certain goal related to the most important quality attributes of the system and the risks that we want to avoid during the development process. It features ways to facilitate critical thinking using Socrative questioning. To check the impact of this game, we have evaluated a game in featuring the testing concepts of the game with computer science students of the NHL Stenden University of Applied Sciences and the Open Universiteit during the second semester of 2023. Results show that students perceived an improvement in the quality of their testing strategies.

Keywords:

Software Testing Education, Gamification, Empiricism.