Hello,
I'm Dimitri Durieux from Cetic, research center dedicated to TIC in
Belgium. I perform risk oriented software analysis using mainly Cast
and SonarSource tools. I worked on the development of CHOOSE
platform, an implementation of the qualoss method to assessment
maturity of the OW2 open-source ecosystem.
My feedback on the scales is the following.
The idea to set the score according to a benchmark makes the whole
projects scores tend to similar ones. The objective of the maturity
assessment initiative you're doing is to improve the
maturity/quality of the projects. Some the scale must not depends
only on the maturity of the actual projects.
Moreover, outliers must be taken into account in the scaling system
mainly if the score is very great.
A solution is to define the maximum (5) to be impossible to reach
for every attribute in your quality model. The scaling number can
have decimals number.
With this system, you ensure that you can distinguish a score of 3.1
and 3.4 for sample. It is very important to motivate the people to
continue to improve the maturity. Every little effort spent on the
project to improve quality or maturity can be clearly identified by
everybody. It will reinforce the model acceptation.
my 2 cents,
Dimitri
--
Dimitri DURIEUX |
Research Engineer - Software &
System Engineering Dpt. |
CETIC
Rue des Frères Wright, 29/3
B-6041 Charleroi
Phone: +32 71 490 767 Mob: +32 494 928 344 |
|
|