Empirical Evaluation of Agile Practices Impact on Team Productivity

Agile methods have become more popular since the early 2000s and, in some cases, can offer better results for software development projects when compared to traditional approaches. Agile methods promise to achieve high productivity and to deliver high-quality software, attracting the attention of companies, which demand ever-higher development speed and quality in their products. The general topic of this PhD thesis is to perform an empirical evaluation of the impact of agile practices on team productivity by means of empirical studies. Regarding the current state of theory and research, more empirical evidence of effectiveness on practices recommended by agile methods are required [1]. Particularly, there are few scientific studies about the impact of agile practices on productivity. In the systematic review conducted by Dyb̊a and Dingsoyr [1], studies just analyse the impact of agile methods on team productivity in terms of lines of code (LOC). However, software development is knowledge work which nature is more complex and harder to evaluate [2]. Measuring productivity with a richer set of metrics would help to deal with this drawback, allowing deeper analysis of the results obtained in the studies. Moreover, agile teams not always collect enough data for a consistent measurement. Therefore, we have to identify the most viable way that best fits the agile spirit in order to measure productivity. Some recent studies discuss the main software development productivity factors [3]. All include the development method as a factor, but in a very superficial way. None of them discusses how agile methods can influence team productivity. In addition, according to Petersen [4], many studies have been conducted before the year 2000. Thus, productivity factors “need to be re-evaluated in currently operating development organizations”.