Empirical Methods in Software Engineering (01OPJIU)

Corso di III livello - Anno 2011

News

• Warning: currently this page hosts the 2011 edition materials
• Course starts on November 22
• Location of lectures: Aula C In the basement, follow directions for Aula 14

Schedule

1. Tue Nov 22 - Introduction, Overview of Empirical Methods in Software Engineering
2. Fri Nov 25 - The experimental process
3. Tue Nov 29 - Software Measurement
4. Fri Dec 2 - Data Visualization
5. Tue Dec 6 - Data Analysis
6. Tue Dec 13 - Surveys
7. Fri Dec 16 - Systematic Literature Reviews (Guest lecture: Prof. Marcela Genero from UCLM)

Material

• Introduction: slides
• Overview: slides
  1. Screencast part 1
  2. Screencast part 2
• Experimental process: slides
  1. Screencast part 1
  2. Screencast part 2
  3. Screencast part 3
  4. Screencast part 4
  • Sample worksheet for the Binomial test
• Software measurement: slides
  1. Screencast part 1
  2. Screencast part 2
• Data visualization: slides
  1. Screencast part 1
  2. Screencast part 2
  3. Screencast part 3
• Data analysis: slides
  1. Screencast part 1
  2. Screencast part 2
  3. Screencast part 3
  4. Screencast part 4
  5. Screencast part 5
  • Paper:
    Ricca F., Di Penta M., Torchiano M., Tonella P., Ceccato M., Visaggio C. A. "Are Fit Tables Really Talking? A Series of Experiments to Understand whether Fit Tables are Useful during Evolution Tasks" IEEE International Conference on Software Engineering, Leipzig May 10-18, pp. 261-271, 2008
  • Raw data
  • Sample session.R
• Surveys: slides
  1. Screencast
• Systematic Literature Review (guest lecture by Marcela Genero): slides
  • example

Goals

• desing, plan, and conduct experimentation (and empirical validation in general) mainly in the context of software engineering, but also in other disciplines of Computer Science;
• perform data collection from all the different sources available during all the range of the software development lifecycle: text, software repositories, human subject observation (developers);
• build statistical models based on the collected data from software products, software processes and experiments

Exam

1. Participate in an experiment first as subject and then as analyst
   • Task presentation lecture:
     1. Introduction to Unimod and MDD
     2. Definition of a model
     3. Performing change tasks
   • Unimod site
   • Task description
2. One of the following two assignments
   • Perform a systematic Literature Review on a topic of your choice
     • The topic has to be agreed upon with the teacher
     • The review protocol must be discussed and accepted before conducting the data collection
   • Select, in agreement with the teacher, an empirical study in Software Engineering and analyze it in detail to identify strenghts and pitfalls