Test Data Generation Approaches for Structural Testing and Automatic Programming Assessment: A Systematic Literature Review

Automatic Programming Assessment (or APA) has commonly been applied in computer science education in assisting educators to effectively mark and grade students’ programming assignments, and hence significantly reducing the workload of educators. APA is one of the methods that require test data generation as its integral part. This study attempts to reveal gaps between software testing and computer science education researches in terms of test data generation for structural testing. This study utilizes a technique of Systematic Literature Review (SLR) to provide a comprehensive review of the existing evidence within the topic of interest. This SLR aims to identify possible approaches to improving the current study of APA, particularly in generating test data for structural testing. The results of this study revealed that there is a clear gap within the perspective context of the covered areas. Particularly, in software testing, the findings show that Meta-Heuristic techniques are among the popular techniques that were applied to realize Automated Test Data Generation (ATDG) for structural testing. Apart from that, the path coverage, branch coverage and statement coverage have been identified as the most popular structural logic coverage that were applied in previous studies.