Understanding acceptance of software metrics - A developer perspective

Software metrics are measures of software products and processes. Metrics are widely used by software organizations to help manage projects, improve product quality and increase efficiency of the software development process. However, metrics programs tend to have a high failure rate in organizations, and developer pushback is one of the sources of problems. Metrics are not welcomed by developers for a variety of reasons including the fear that metrics will be used to monitor performance, distrust of the metrics process and a lack of belief that metrics will be useful. This research study had two objectives - to identify factors affecting acceptance of software metrics in organizations, and to evaluate the usefulness of our approach in explaining acceptance of a metrics program in an organizational context. To address these objectives, we created a diagnostic survey tool called the Metrics Acceptance Model (MAM) that gathers opinions and attitudes about the metrics program directly from developers. The survey, and underlying framework, is based on problem areas identified in metrics research and technology acceptance literature. To evaluate the association of these factors with developer motivation to participate in metrics programs, we applied our survey to two different samples of metrics practitioners, revising the survey and the model in between. We learned that two factors: metrics usefulness and attitude towards metrics were significantly related to the motivation to collect metrics data. Further, we conducted a case study to evaluate the usefulness of the MAM survey tool. We used a mixed-methods approach in which we conducted interviews and surveys of project managers and developers involved in a metrics implementation. Our results show that the MAM was able to provide useful insights to project managers about developers' lack of confidence in metrics and lack of control over the metrics process. We also found that the MAM can be an effective diagnostic tool both at an organizational and project level to identify potential impediments in metrics programs. Our research highlights the importance of the developer perspective in a software metrics initiative with emphasis on issues such as fear of adverse consequences, managerial influence, and perceived capability of contributing to metrics. To our knowledge, a systematic analysis of such factors has not yet been addressed