I'm a software manger that is leading a team being appraised for CMMI. Today one of my developers asked my why we needed to produce an "indirect" piece of evidence for each practice in the appraisal. We produced the direct for our Lead Appraiser - what, don't you trust us?
Ahhhh…. I love this question! No worries – many incredibly smart people struggle with the logic of this one, so you’re in good company! And yea, we trust you . . .trust by verify comrade.....
I like to say that the “Directs” are the least important thing in the model, with indirect evidence and the GPs being the most important things.
The indirect evidence (or affirmation) are the only thing that indicates that a process was actually followed to produce the end result (the “Direct”).
For instance, let’s say an ATM is reviewing an estimating worksheet. It has all the attributes required of the direct. It has sizing, counts of requirements, test, functions, etc. It takes into consideration all of the phases of the lifecycle, it was approved, and signed off on. All in all, one FINE direct. A chip off the ‘ol “CMMI block." Fully Implemented, right? Whooooaaa! Hold on a second there big fella!
Did some developer just sit at her desk and plug these numbers in? Did the approver actually look at it before their admin signed it with his signature stamp? Was it peer reviewed? Hell, is this even REALLY an estimate from the project being appraised?
Hmmmmm, vewwwy twicky qwestions. We need some other method to increase the odds that we’re right when we make all those assumptions and lay a big, fat “FI” on them. Ah ha! Bring on the indirects!
That’s what I would say . . just like that J
www.broadswordsolutions.com
