Case Study Data to Evaluate the ASPLe Methodology

Published: 1 May 2020| Version 1 | DOI: 10.17632/y7b5pdtv9k.1


The dataset archives a case study that was conducted to evaluate a novel Autonomic Software Product Lines engineering (ASPLe) methodology. ASPLe provides process support to design and develop product lines of self-adaptive systems with systematic reuse. The case study evaluates ASPLe by comparing it with a reference approach. Monitor-Analyze-Plan-Execute-Knowledge (MAPE-K) feedback loop was used as a reference approach. The comparison was performed using test assignments, questionnaires, and interviews methods. Test assignments comprised tasks to design self-adaptive systems using ASPLe and the MAPE-K reference approach. Subjects of the case study were students of a final-year master's program in software engineering. We divided subjects into two groups according to block subject-object study classification to do the test assignments. After test assignments, we required subjects to compare ASPLe and the MAPE-K reference approach mainly with respect to their support for software reuse and uncertainty mitigation. Subjects were also interviewed to clarify questionnaire responses and collect additional details. Interviews were done in a semi-structured way. Data collected from tests assignments, questionnaires, and interviews were analyzed for the main objective of the study. The main objective of the case study was to compare ASPLe with the MAPEK reference approach for design support to maximize software reuse and mitigate uncertainties. We hypothesized that ASPLe provides better support for software reuse and uncertainty mitigation than the MAPE-K approach. The case study data were analyzed using different analysis methods. Tests assignments data were analyzed using Shapiro-Wilk, Paired-Samples T-Test, and Wilcoxon statistical tests. Questionnaire data were analyzed using descriptive statistics and graphs. Interview data were analyzed using qualitative content analysis method. The results of the analysis showed statistically significant results in favor of ASPLe. In particular, the results showed that ASPLe helps developers to increase reuse and mitigate uncertainties in the development of self-adaptive systems with reuse.



Uncertainty, Software Engineering, Autonomic Computing, Software Design, Software Process, Reusable Software, Domain Engineering, Methodology