In practice software testing can take half of total time needed for development, which means that choosing the right test methods is very important. The performance of selected test methods is nevertheless very rarely evaluated. Most models on the market evaluate the performance of the model based on objective evaluations of respondents and do not take the differences between process managers and users of the process into account. In this master's thesis we devised a model for the evaluation of test methods, which takes both views into account: technical suitability and acceptance among users. The model was first conceived in theory with verified evaluation methods and six different dimensions. Theoretically based model was then verified in a company that specializes in the development of mobile applications. Three teams evaluated test elements they use and we analysed the results. Team leaders have confirmed that with our model they have received useful information.
|