Some Perspectives on Teaching Experimental Software Engineering
Producing software products in the large is a problem-prone activity where projects turn too often in unsuccessful stories. Software Engineering (SE) is the knowledge area that merges contributions from the Engineering, Management and Quality fields to propose solutions to the above mentioned problems.<P>
In recent years SE has been evolving from a traditional qualitative posture to a more quantitative one, which makes it closer to more mature engineering fields. This quantitative (r)evolution in due to the emerging field of Experimental Software Engineering (ESE). The objective of ESE is primarily to understand the cause-effect relationships in the software development process by the use of the scientific method: raising hypothesis and trying to corroborate or refute them by means of experimentation. For that purpose ESE uses applied statistical techniques and is related to SE in a similar way as the field of Econometrics is related to Economics.<P>
Several universities throughout the world have started to include courses on ESE in their Computer Science curricula. We will start such a course, this school year, both at the undergraduate and post-graduate levels. Due to the new problems and solutions that we plan to expose our students to, we would like to share some perspectives on both the theoretical and the practical aspects of teaching this emerging area. To frame the topics dealt with in ESE we will present a brief overview of some case studies, taken from our research in this field since the mid-nineties, that lead to the creation of the QUASAR ( Quantitative Approaches in Software engineering And Reengineering) research group.
