“Software Engineering Experimentation” tutorial
Mar 21-25, 2011 (20hs)
Dictated by Natalia Juristo at Universidad ORT Uruguay
Content: This seminar aims to understand and improve the use of evidence based methods in Software Engineering Research. The focus is learn about the design and analysis of controlled experiments in order to compare different software engineering techniques. Most of the widespread knowledge on building software systems is based on proposals from authors, theoretical deductions or proofs of concept with no empirical basis formalized. In contrast, in other scientifics fields or engineering branches, experimentation is the usual way to support the generated knowledge. This course intends to set the relationship of experimentation with a theoretical framework for generalizing results and contribute to the body of knowledge of this area.
It was dictated by Dr. Natalia Juristo, PhD in Computer Science at Universidad Politécnica de Madrid and Martín Solari, Associate Professor of Software Engineering Department of the Universidad ORT Uruguay.
- Introduction to empirical methods.
- Status of empirical research in software engineering.
- Case studies and surveys methods.
- Experimental concepts: hypothesis, factors, parameters.
- Design of controlled experiments.
- Methods of statistical analysis.
- Replication of exeperiments
- Construction laboratory packet.
- Aggregation of results and meta-analysis.
- Relationship between theory and generalization.
Bibliography
- Juristo, N., Moreno, A. 2001. Basics of Software Engineering Experimentation. Springer.
- Wohlin, Claes, et al. 1999. Experimentation in Software Engineering: An Introduction. Springer.
- Antony, J. 2005. Design of Experiments for Engineers and Scientists. Butterworth-Heinemann
- Shull, F. et al. 2007. Guide to Advanced Empirical Software Engineering. Springer.
