converter writes "The Brainbench Perl 5.8 Exam Beta is now available online (click the 'Get Test' button to proceed). I was a paid reviewer during the development process and submitted a number of corrections with working code samples and documentation. If you would like to add your input, please take the free beta exam and submit feedback." This was submitted a week or two ago; I don't know if it is still in beta.
"Our Perl 5.8 test measures your knowledge of programming in Practical Extraction and Reporting Language."
Re:Practical Extraction and Reporting Language
gabor on 2005-08-27T05:03:10
s/kus/just/
of course
Another concern is that a single test should not try to test more than is feasible, or necessary. All existing Perl certification tests seem to view Perl knowledge as unidimensional, and Perl programmers as homogenous entities, amenable to assessment through a single test. A more realistic view is that some Perl programmers (e.g., those working in System Administration roles) may need only the most rudimentary kind of "procedural" Perl programming skills, while others (e.g., those working on large projects with other developers) may well need more advanced knowledge, including familiarity with OO techniques.
I believe that:
Interested JAPHs have engaged in spirited discussions about fashioning certification tests that cater to the unique properties of Perl and its JAPHs. Due to the effects of book interference on my supply of spare time, progress on this project has been on hold for some time, but I plan to resume work on it soon. I encourage those who want to help, in roles ranging from surveying IT managers to writing the actual certification tests, to sign up on this web page.
Re:BrainBench
yumpy on 2005-08-29T05:02:02
We've already had some discussion of whether Larry would have to endorse a certification suite to make it work; see this page; patches welcome!Re:BrainBench
KM on 2005-08-29T13:52:56
That's not really a discussion, that's you writing something:-) Larry doesn't _need_ to endorse it, but someone does, in my opinion. If there isn't an endorsing body, then anyone can make a "certification" test and nobody will know what the validity of those tests are.
What would stop me from making the "Meltzer Super Duper Perl Certification Exam", charging people $20 to take it, giving them a certificate with a gold star and my signature? Nothing. And, what would make that exam less valid than BrainBench or any other exam?Re:BrainBench
yumpy on 2005-08-29T20:17:32
"The proof of the pudding is in the tasting."If your test really helped hiring managers separate the JAPHly wheat from the chaff, word would get around, and your test would acquire a favorable reputation, and possibly become a de facto standard. Which is as it should be; the users of a certification service should determine the success or failure of a certification suite.
Consider the chain called "Seattle's Best Frozen Yogurt" -- their name sounds very impressive, but I'll take Ben & Jerry's product over theirs any time, because, despite their lame name (say that 3 times fast!), the latter has a sterling reputation that's well deserved.
Just as Apache, Ebay, and Google became famous for providing valuable services reliably, without any "celebrity endorsements", a Perl certification service could do the same.
But don't get me wrong--I'm not saying that it wouldn't be advantageous for Larry, or Damian, or Ingy, or perhaps even Spongebob to endorse a fledgeling Perl Certification suite, but that might not be realistic to expect.
On the other hand, don't you think the PR buzz for a Perl Certification suite developed by actual working JAPHs, for the benefit of their community, would have a lot of appeal to corporate hiring types -- who are used to relying instead on large corporations with vested business and financial interests in flogging a particular technology?
Re:BrainBench
KM on 2005-08-29T21:14:13
Let me say I haven't read through your whole site. I know I did after we met when I hosted YAPC. I'll need to see if my ideas mesh with yours.. feel free to point me to places in your wiki where I may need to read.First real question is if certs are needed at all. Who asks for a Perl certification out of a perspective employee? I'm also not really sure what a cert would do. I may have missed it, but what I really see from your effort is similar to BrainBench. You take a test, you pass, you get a certificate. That doesn't make you certified, it means you passed an exam. There is no classroom time, not required reading, no practicle, no labs, etc... As well, there is no approving body of the exam aside a group of people who made it and say it's good.
So, what real use is that? What exactly is the demand for that? How do you know I don't then charge people more money to take your exam for them? (pause for thinking I may be talking myself out of a money making opportunity here).
If your test really helped hiring managers separate the JAPHly wheat from the chaff, word would get around, and your test would acquire a favorable reputation, and possibly become a de facto standard. Which is as it should be; the users of a certification service should determine the success or failure of a certification suite.
What's the suite? How does a PHB know that your test is better than BrainBench? Is it better? Is the test for certification pointed towards what the company needs are?
But don't get me wrong--I'm not saying that it wouldn't be advantageous for Larry, or Damian, or Ingy, or perhaps even Spongebob to endorse a fledgeling Perl Certification suite, but that might not be realistic to expect.
If Spongebob gave his OK, my daughter would be happy. I don't think any respected person who would be wanted to endorse it, would. Just as TPF has been staying out of the discussion.
On the other hand, don't you think the PR buzz for a Perl Certification suite developed by actual working JAPHs, for the benefit of their community, would have a lot of appeal to corporate hiring types -- who are used to relying instead on large corporations with vested business and financial interests in flogging a particular technology?
Not sure. It depends on who the JAPHs are, and if their combined name-weight carries PR buzz. But, I'm sure that wouldn't be hard to do.
I think what bothers me is the "certificate"/"certified" term. No single test, or couple tests, can really certify that someone is a Perl Programmer. I think it would be more useful to just come up with a way to quickly tailor Perl Pre-Employment Exams. That way, Foo Corp. can use this well-respected coven of JAPHs to make an exam which leans to what they do/need out of an employee. Foo Corp. may do bio work, and cares more about certain things than Bar Corp. which does shared web hosting, to Zog Corp. which does e-commerce. I think HR people would love this. Nobody would be 'certified', but these tests could be used during the pre-employment phase by companies. As well, a set of well-rounded tests could be made which can help people guage what their skill level is. This could help encourage people to keep learning. If they pass Level 1, great.. they move on to Level 2. But, they aren't "certified".. rather they completed a 'stage'.
The only certificate I can see existing is that of training companies. If I take a course, I get a certificate saying so, and if I pass a final exam based on that course, I get a certificate saying so (Certified Completing XYZ Course). Not only did I take the course, I absorbed something from it. Or, there needs to be a certificate program. Similar to MCSE, A+, etc... You go to classes to prepare for an exam which gives you a certificate. It's a process, not a test, and I can be certified that I completed the process (still, not saying I am a certified Perl Programmer... as that will really be deciced on the job).
So, tell me it's a service to provide companies with pre-employment tests which meet the needs of the company and guage a person based on that skillset.. I'm in.
Tell me it's a progression of tests and materials to help a programmer become a better programmer.. I'm in.
Tell me someone gets to be a "Certified Perl Programmer".. I'm out.
Re:BrainBench
yumpy on 2005-08-29T23:47:03
Many of your comments are along the lines of FAOs ("Frequently Asserted Objections") to certifications in general, which I've addressed in my TPJ article entitled Is it Time for Perl Certification? , and others are addressed in discussions on the Perl Certification Wiki.Regarding a few of the other points you raised:
- Certification needn't be a monolithic, PASS/FAIL thingy; it can be multi-level, with job-specific and diagnostic elements too
- a Certification Test needn't be a 100% multiple-guess trivia test; it could also state program specifications and ask for on-line source code input, for "grading" by Perl itself, and human reviewers.
- it's important to ensure that any career-influencing test be psychometrically reliable and valid, and logistically secure--the current on-line tests don't meet these standards--ours could!
- regarding the link you want to draw between Training and Certification, that sounds like a way to bolster the prestige of a test having dubious validity by pairing it with actual exposure to the subject matter. A psychometrically valid test wouldn't need this kind of crutch, because we wouldn't care where the testee picked up the knowledge, so long as we know he really has it
- your idea of job-specific tests has come up before, and has a lot going for it. Somewhere down the line, I'd like to see that happen, but first, we'd probably have to restrict our focus to testing the basics
We all spent a couple of years trying to get off brainbench's mailing-lists. I hope their spamming ways have improved since then, but no way am I giving them an actual email-address again.
