Case studies that show the effectiveness (or not) of Agile Software Development. Please provide a brief summary of each entry.
Contents
As a whole
XP Productivity Numbers (2004) by Jeff Grigg.
This isn't pulled together as a paper, but Jeff Grigg made a spreadsheet out of numbers reported by Mike Corum on some XP projects and calculated the time and cost savings realized with XP. See http://tech.groups.yahoo.com/group/xpstl/message/1193 for the description. Find the spreadsheet "XP Productivity Numbers 2.xls" in the Files directory http://tech.groups.yahoo.com/group/xpstl/files/
The Total Economic Impact™ of Using ThoughtWorks' Agile Development Approach (2004) by Forrester Research, Inc. http://www.thoughtworks.com/pdfs/forrester_tei.pdf
In August of 2003, ThoughtWorks commissioned Forrester Research to examine the Total Economic Impact™ (TEI) of an organization engaging with ThoughtWorks for a software development project using an agile approach. The organizations interviewed by Forrester all showed actual or expected returns on their investments in ThoughtWorks that ranged from 29 percent to 64 percent on a risk-adjusted basis and from 31 percent to 66 percent on a non risk-adjusted basis.
Alistair Cockburn's Doctoral Thesis (2003) http://alistair.cockburn.us/index.php/People_and_methodologies_in_software_development
Rather than setting out to prove the effectiveness of an agile process which had already been defined, this work set out to identify effective processes by studying real teams in action. Alistair Cockburn documented his findings in the form of the "Crystal" family of agile methodologies. In this case, the study created the agile methodology (and thus, clearly, supports it!)
Agile Practices
Other
Predicting Software Defect Density: A Case Study on Automated Static Code Analysis (2007) Artem Marchenko (Nokia), Pekka Abrahamson (VTT Technical Research Centre) http://www.springerlink.com/content/n31723r4268j3054/?p=60711d7593c148b6ada32e1c772a39f8&pi=17
- Results of a study of static code analysis as a means to predict defect density in a software product.