• Add a Comment
  • Edit
  • More Actions v
  • Quarantine this Entry

Comments (1)

1 mcantor@us.ibm.com commented Permalink

This morning Sateesh Kannegala referred me to a paper by Cern Kaner and Walter P. Bond ( http://www.kaner.com/pdfs/metrics2004.pdf ) based on the IEEE Standard 1061, a framework for creating a measurement framework for software attributes.

This standard has not gotten much traction. I suspect their are several reasons for the lack of acceptance, chief among them is that it is not based on given process discipline such as SPC. Instead they define direct and indirect measures. As the authors of the paper point out, the standard defines mean time to failure (MTTF) as a direct measure. Kaner and Bond point out treaing MTTF as a direct measure is not useful. That is consistent with this post. In our approach MTTF is an outcome we want to control. The controls and control measures must be based on team practices.

Add a Comment Add a Comment