Agile Teams
Publications - TSP-EF Presentation at TUG 2004
(Second Annual TSPSM Users Group Conference)
"TSP-EF Proposal: A Team Software ProcessSM Evaluation Framework"
Authors | Publication History | Related Materials | Background Required | Abstract

 
Authors:
This paper was co-authored by Karen Smiley, Jan Höglund, and . See the final slides for acknowledgements and credits to contributors.
 
Publication History:
This presentation is the first in a series of papers we are writing on TSP-EF as it develops.
  • August 27, 2004: , slides submitted to SEI - comments invited.
  • Sept. 27-28, 2004: V0.7 slides presented at TUG by Jan.
    (TUG 2004 Proceedings papers are no longer available on the SEI website at )
 
Related Materials:
Background Required:
This paper assumes good familiarity with the and . It will be most meaningful to experienced TSP practitioners or coaches. No prior knowledge or experience with XP-EF or other Evaluation Frameworks is required.
 
Abstract:
(After the conference, the abstract will also be available on the ABB Corporate Research Scientific Publications website)
 
This presentation will describe to conference attendees a proposal and first draft of an approach for measuring how closely a Team Software Process project team is following all of the TSP and PSP principles and practices. The proposed TSP Evaluation Framework (TSP-EF) would supplement, not replace, existing TSP coaching tools such as the Checkpoint script. It parallels, and builds upon, the three-part Extreme Programming Evaluation Framework (XP-EF) being developed by Williams, et. al. in the agile world for measuring XP project practices and results in industry.
 
The goals of developing TSP-EF are two-fold. In the short term, it can provide a quantitative, consistent method for TSP coaches to measure which practices and principles of the Team Software Process and Personal Software Process are actually being applied on their real-world projects, and to what extent. In the long term, it could establish a foundation for future correlation of results vs. practices, enabling statistical analysis of which practice factors are most influential in achieving the high quality and improved predictability which TSP and PSP (properly used) can deliver.
 
The purpose of this session is to present an initial straw-man TSP evaluation framework, and elicit feedback from TUG attendees on:
- whether such a framework would be valuable to them;
- which context, adherence, and results factors should be included, or excluded, in the framework;
- how each element of the framework should be measured;
- how to make it easy for TSP coaches and teams to effectively apply the framework to guide continuous process improvement on their projects.