Tuesday, October 18, 2011

Please, no new "certifications"

I just saw an advertisement for this Building a Certification Testing Program - Cutting through the hype to see how it really works on LinkedIn, and I couldn't stop myself from adding the following comment:
Please make it stop. We don't need more "certification" programs -- not unless you are going to be the first organization that allows itself to be held legally and financially accountable when people you "certify" can't do what you "certified" they can.

Otherwise, conduct all the training you want. Assess student performance if you want. Only "pass" students who "pass" the assessment if you want.

Just do us all a favor and *STOP* calling it certification until you are willing to do things like:
  • reimburse hiring expenses to employers who hire folks you certified as being able to X who can't X
  • implement periodic re-assessment to enforce some bar of continued knowledge/skill/ability over time
  • implement some way to revoke certifications of folks who fail to demonstrate knowledge/skill/ability in the workforce
The list goes on, but I know it's pointless. The certification machine will continue no matter how loudly, or how frequently I point out the ways in which it is frequently (at least arguably) unethical and fraudulent - at least in "testerland."
Seriously, this drives me insane.  Others can make stands about content, assessment methods, etc. -- I have my opinions on those things, but honestly that part of the topic bores me.  People decide what university to attend, what to major in, what electives to take, etc. for their degree programs ... they can decide on whether or not the content of some professional development program (with or without "certification" program) is worth their effort.  What I want to see is the "certifying bodies" being held accountable for complying with the claims they make about the individuals they "certify."

I mean, seriously, have any of you seen any data that you'd consider either statistically significant, empirical (vs. anecdotal), or free enough from obvious experimental design flaws to support the claims we see from "certifying bodies"?  If you have, please share the data with me and I'll list it in line -- unless of course, it's flawed, in which case, I'd be happy to point out how and why the data doesn't support the conclusion.

Otherwise, please, please, please don't engage in creating more of these things.  Please.

Scott Barber
Chief Technologist, PerfTestPlus, Inc.

Co-Author, Performance Testing Guidance for Web Applications
Author, Web Load Testing for Dummies
Contributing Author, Beautiful Testing, and How To Reduce the Cost of Testing

"If you can see it in your mind...
     you will find it in your life."


Kiril Varbanov said...

Actually Scott, the more you talk about "certification" and some of its flaws, the more the companies will start paying attention to "certified" people, I mean, they will know, at some point, that this "certification" doesn't mean a lot.

So, in an ideal world, the real skills and consistency will make sense to an employer, rather than the "title".

Anonymous said...

Actually I feel Scott, you are reading this webinar somewhere wrong, there is new term which has come into testing world."Certification Testing" which I am coming across from past 2 years or so and this webinar has nothing to do with individual certification offered to individuals.Appslabs as far I know is not into individual certification.

Scott Barber said...

Fair enough, I fully admit to not having read the advert closely (honestly, the title "flipped my bozo bit" before I had a chance to consider the details).

That said, the this post accurately represents my reaction to the title and makes points that I believe in strongly -- whether or not they apply to this webinar.

So my apologies to AppLabs if I have mischaracterized or offended.

Matthew Sullivan said...

Perhaps part of the problem is the inability of the management class to continue to understand the complexity of what their underlings are doing. We have "resource managers" hiring .NET engineers without really knowing anything about .NET. So they look for people "cerified" by Microsoft assuming that means they know .NET. In order to sustain that model, Microsoft continually makes the tests more difficult and less relevant to real-world engineering in the age of searchable knowledge (ie Google) who have memorized key bits of code without really knowing how to use them to solve human problems.

Similarly with testing certifications, but much worse. Since the perspective of most testing certifications seems to be that testing is more of a business activity than a scientific approach to seeking knowledge about a subject, the material memorized by certified testers is more about process than testing, and is worse than useless. If implemented, it can actually comprimise the real work of testing by requiring unnecessary distractions.

The solution is for businesses to evaluate and reward based on accomplishment, not metrics or certifications. Unfortunately, this is something that requires understanding what the people who report to you are doing. In other words, we should be working towards a healthy promotion path, and not headhunting Devry MBAs to run our testing departments.

Meanwhile, I will continue to pursue certifications such as Microsoft's if I feel like I can learn something useful from the exercise. When I becaome an MCP in Enterprise Application Development, I don't expect to really know how to build enterprise applications. But I do expect to know a little a bit about what they do.