Track: Graduate Student Paper Competition

Abstract

Software testing effort estimation is a critical phase in software project management. Software testing is an integral part of the software development lifecycle (SDLC). It ensures that requirement specifications are met before a product is deployed. Accurate test estimation is crucial to the on-time delivery of a project and ensures that it meets resource constraints. In this study fuzzy logic is used to estimate the manual test script authoring effort in the test process. MATLAB software is used to design the fuzzy logic system that would calculate the duration of the scripting effort. The variables considered were test complexity, test script reusability, and number of test scenarios. The crisp output is the effort estimation in hours. The Delphi method was used to collate the ratings of each expert for the parameters of each membership function in each variable. The linguistic rules were defined after. Crisp inputs for each project were applied to the model and the output was compared with the actual effort duration of previous projects and expert-based estimation. The results showed that the mean absolutive relative error (MARE) of the fuzzy logic model has performed better than the expert-based estimation technique. Thus, the fuzzy logic framework may be used as a decision support tool in manual scripting effort estimation. This study is the first to explore the effort estimation in manual test script authoring. This model could serve as a baseline to test engineers, test leads, project managers, and business analysts.

Keywords

Software testing, Effort Estimation, Software Supply Chain, Manual Testing, Script Authoring