An analysis of parameters influencing test suite effectiveness
MetadataShow full item record
Software Testing is one of the most critical areas in the Software Development Life Cycle. Software Testing is quite expensive to implement, and it does not guarantee good results, unless it is implemented perfectly. If better testing were performed, it would ensure that a huge cost is not spent in reporting undetected software bugs. For enhanced testing to be achieved, it is extremely vital to understand how to build an effective test suite, which would test the software system in the best possible way. An effective test suite needs to be built for better testing to be implemented for any product or service. In this study, we try to understand how to build an effective test suite by analyzing various factors which influence its effectiveness like test suite size and coverage level. Though size and coverage are the most common factors, we would also like to determine if there are any other factors, which also have an impact on the test suite effectiveness. We try to come up with a relationship between the factors like coverage level, test suite size and effectiveness. A point to note here is that most of the factors influencing the effectiveness may be covariate. So, it is of vital importance that we keep one factor constant in order to measure the impact of the other factor. This helps us to understand that which among these factors is most crucial to the effectiveness and by what degree is the impact of that factor on the effectiveness of the test suite. Effectiveness here is measured by fault finding effectiveness of the test suite, i.e. the number of faults (errors) that can be detected by the test suite. Improving the effectiveness will improve the testing, and thus contribute vitally to the final outcome of the work. In this thesis, we propose and evaluate an estimation through which effectiveness can be calculated in terms of the factors influencing it, like coverage and size.