The Certification Debate
January 06, 2007
The pro’s and con’s of certification in Agile Leadership are being discussed on a few Yahoo groups at the moment. Some people believe any kind of certification is inherently evil and I can appreciate this point of view, but the idealist in me wonders if it can be done right?
I am involved in an APLN sub-group to create a proposal for an accreditation scheme. I thought long and hard before joining the group, but it became apparent that other groups and training companies were discussing the creation of alternative schemes and so my feelings were that if it was going to happen anyway, then I would like the chance to steer it towards a decent goal.
The current situation
It appears that people are eager for direction and recognition in agile project management. We can choose to step aside, avoid the nasty politics, ego’s, monetary issues, and bureaucracy that often go with it and risk witnessing other groups implement poorly directed schemes. Or, try to create what an accreditation (I’m avoiding the term “certification”) scheme should be and steer the industry towards, what we believe, are the right goals.
My certification observations
Like many others, I have been through the ScrumMaster, PMI, and DSDM accreditation schemes and observed the following: ...
ScrumMaster Certification – this is a very popular certificate for attending a two day training course. I sat an early class with Ken Schwaber and recall it being an entertaining and educational course.
Verdict: cheap to administer, however it only really proves that someone has attended a two day course.
PMI PMP – Requires 4500 hours of on the job experience and 35 hours of PM related training. You then sit a 3hour multiple choice exam with a fairly low passing grade.
Verdict: Tests for conformance to a somewhat wrong-paradigmed set of ideas. Provides some measure of temporary recollection (of somewhat wrong-paradigmed set of ideas) and professional experience. The testing is computerized multiple-choice which is cheap to administer once created.
DSDM Practitioner – (I took this 8 years ago, it may have changed since) required 1 years experience on DSDM projects. You then submit a 10,000-15,000 word account of the project and later go for a face-to-face interview with a DSDM Examiner who asks you about the DSDM principles, their practical use and limitations, your project experience, etc. It had a high failure rate but, re-tests were cheap.
Verdict: It felt like you actually had to understand and know how to use DSDM to pass. The side affect of requiring project accounts is that very quickly the DSDM Consortium built up a valuable library of real project case studies (that they did not fully capitalize upon due to not asking permission to re-use them). However, the whole process was very labour intensive, yet self-funded by exam fees.
I think agile project management accreditation could work if it had the right guidelines. My suggestions would be:
1) Create a name that indicates that the holder has reached the state of Learner not Master. Perhaps something like “Agile Project Management Learner Level” or “Agile Leadership Foundation Certificate”.
2) Get people to put some conscious effort into it by requiring X months experience on real agile projects and writing up an account of a project for review that can also be reused by the APLN and its members as part of a growing library of case studies.
3) Create an approved training curriculum and potentially courseware to be vetted by the agile community, used for training, and updated regularly.
4) Once created by a central working committee, perhaps local agile chapters could administer the training and accreditation process to avoid the need for a centralized administration model (this may be an naïve hope.)
These are my thoughts. It is always easier to find faults in something than create a better alternative, but I think if any group can create an effective accreditation program then it is the wisdom of crowds within the APLN membership.
Please tell me your suggestions. If you were charged with creating an accreditation program, how would you make it the very best it could be? (Please feel free to post your suggestions as comments here, or email them to me at [email protected])
I agree wholeheartedly with you on the status of existing accreditation schemes, having been through the ScrumMaster course and my experience with CIPS' I.S.P. desigation.
To keep my thoughts organized, I'll reply to each suggestion in turn:
1) I quite like the word "foundation" as it evokes the right emotions for me - a solid foundation, a journeyman status, but not master.
2) I like the concept, however what constitutes a "project" would have to be defined. For example, I work as a consultant, doing development work, but also coach my clients in the use of agile principles on small projects. Unless I've misread the goals, I think this type of "professional" experience should be included. To borrow from CIPS' definition, work is considered professional if you are accountable for the results and not simply following someone else's detailed plan.
3) A Body of Knowledge (BOK) that defines what is reasonable knowledge to possess would be a nice addition. Although I'm not suggesting that we redefine Scrum, Lean, XP, etc. in this document, we should attempt to capture the essence of what it is to be agile. Training and accreditation would emerge naturally to meet the goals of the BOK.
4) This model reminds me of Scrum training. You progress from Certified to Practioner to Trainer. A similar model could be applied, allowing for great customization of material to suit the needs of local communities. The benchmark could be conformance to the BOK.
A few last words in a more generic context. For me, for an accrediation to be recognized and desirable (i.e. effective), it must show that the individual has solid experience and truely grasps the fundamentals. One significant component, I believe, is a live interview, similar to a thesis defence. Writing a paper is easy (e.g. look at the number of fluffed up resumes out there), but backing that up in front of others weeds out the imposters rather quickly. I'm not advocating a draconian approach to make this an exclusive club, but having a small panel (2-3) be able to ask questions and dive into some depth guarantees that the candidate understands what they are talking about.
Anyways, that's my (brief) 2 cents worth... I'd be happy to go into more detail, just email (olivier -at- ait -dash- consulting -dot- ca).
Posted by: Olivier | January 08, 2007 at 09:27 PM