Evaluating Approaches to Selecting Design Thinking Techniques : Quantitative and Qualitative Analysis.

Context: Requirements Engineering (RE) is essential to software quality. Studies have shown that software engineers often make mistakes, such as insufficient or misunderstood requirements. Therefore, it is necessary to support all the RE phases, especially eliciting requirements. In this context, Design Thinking (DT) is commonly used to deal with these problems. DT aims at bringing quality to software development to achieve users' needs. It has a set of techniques and methods that can help software engineers properly elicit requirements to achieve this goal. However, in the literature, there are several techniques and selecting an appropriate one is not a trivial task. Some approaches support the selection of techniques, including DTA4RE and Selection Universe. Objective: This paper aims to analyze the performance of these two approaches to selecting DT techniques in terms of accuracy. Method: We conducted a controlled experiment to obtain the data. We have applied quantitative and qualitative analysis to the data. Results: Regarding the quantitative results, we found no significant difference in accuracy between the Universe Selection approaches and the DTA4RE. Regarding the qualitative results, we found that grouping the techniques into categories presented by the Selection Universe allowed us to reduce the search time for the techniques from DT since the approach allowed us to associate the system features with the techniques that can be used. Conclusion: We found that the two approaches satisfactorily supported the selection of DT techniques for various requirements elicitation activities. In the participants' perception, the selection universe was helpful because the approach is intuitive. Also, the categorization of techniques made it easier to find the appropriate techniques for the proposed scenarios.