An empirical evaluation of RAIDE: A semi-automated approach for test smells detection and refactoring

Effective test code refactoring is essential for maintaining the quality and efficiency of software development. Automated support for test code refactoring can significantly enhance its cost-effectiveness. However, there is limited evidence on the effectiveness of such automated support, especially for addressing Assertion Roulette and Duplicate Assert test smells. To address this gap, we present RAIDE, an Eclipse IDE plugin that provides an easy-to-use approach to identifying and refactoring test smells. In this study, we conduct a controlled experiment with twenty participants to evaluate RAIDE's capability to detect and refactor Assertion Roulette and Duplicate Assert test smells. The results demonstrate that RAIDE outperforms comparable state-of-the-art approaches in detecting test smells, and can detect and refactor test smells in a fraction of the time. Our findings highlight the potential benefits of automated test code refactoring support for improving the efficiency and effectiveness of software development, and suggest avenues for future research to extend and evaluate RAIDE for handling other test smells and refactoring techniques.(c) 2023 Elsevier B.V. All rights reserved.