An empirical investigation on the satisfaction levels with the requirements engineering practices: Agile vs. waterfall

Agile development practices have become widely accepted as an effective project management approach in order to have rapid delivery of high-quality software. As with traditional waterfall projects, effective communication is also a necessity for the success of an agile project. However, the agile principles set a different tone for project development. The difference in tone is also reflected on how requirements are captured, analyzed and communicated under the agile umbrella. Little contemporary data exists that document actual practices of software professionals for software Requirements Engineering (RE) activities in agile environments. To remedy this deficiency and provide useful data to other researchers we conducted a survey study on the current RE state of practice. In this paper, we run a series of Mann-Whitney-Wilcoxon (MWW) tests to examine a set of null hypothesis on the performance of and satisfaction with various RE related activities in environments where RE are practiced and communicated in agile context in comparison to the classical waterfall context. The results indicated that there is no difference that can be determined between the agile and waterfall in regards with the satisfaction with the RE practices, but also a higher level of satisfaction for agile sample in comparison to waterfall) can be determined in regards to aspects related to productivity and final product quality.