The impact of agile approaches on software quality attributes an empirical study

Agile software development describes those software systems which undergo rapid changes as a result of the testing and requirements fulfillment processes. This development technique came into view in order to overcome the drawbacks of long software life cycles of traditional development methods. This paper investigates the effects of agile practices on the quality of the produced software systems. We have used 20 open and closed source systems of various sizes and functionalities. While the development process of 9 of the studied systems followed agile approaches, the rest were developed using traditional approaches. Firstly, a set of software metrics is generated to describe each system. The metrics encompass complexity and inheritance characteristics of the studied systems. Secondly, the generated metrics are used as predictors of the type of the followed development process using binary logistic regression. The obtained high goodness-of-fit measures show the strong relationship between the used metrics and the type of the followed development process. More specifically, the study reveals that following agile practices has a great impact on lack of cohesion of methods, fan in and maximum depth of inheritance tree.