Uma proposta de métricas para avaliar a efetividade da execução de testes de software

Abstract

The market demands high-quality software, which is delivered on time at the agreed cost. One of the main software industries’ worries is the assurance of the created product quality that has generated the establishment of the software development associated with the main concepts of software quality. Many companies have been investing in the testing process to prevent and identify defects. Software testing is an important ally in quality assurance, which justifies a portion of the development costs being related to testing activities. During the development lifecycle and the software maintenance, tests are executed with the purpose of ensuring that the number of defects is minimized before the final product delivers to the client. The purpose of this research is to develop an approach to evaluate the execution of software test. To achieve this goal, it was used the Goal-Question-Metric approach, which seeks to generate a set of metrics according to the objectives defined by the need of the situation. After the definition of the objectives to evaluate a software test execution, it was proposed the adoption of a set of metrics to facilitate the monitoring and improvement of the software tests execution. To validate the set of proposed metrics, a focus group was conducted with specialists in the area of software testing. Thus, this research offers contributions to the metrics used to evaluate the execution of software tests that currently means a piece with a high cost for software companies.