Beefy Boxes and Bandwidth Generously Provided by pair Networks
good chemistry is complicated,
and a little bit messy -LW

Some reflections on the Brainbench Perl Test

by metaperl (Curate)
on Dec 12, 2008 at 22:50 UTC ( #730066=perlmeditation: print w/replies, xml ) Need Help??

I took the Brainbench Perl certification test as part of a job hiring process. I will start with some objective facts about the test:


There is not a single question on this test about object-oriented programming.

no training materials

There is no in-person or written materials which they provide as preparation for the test. Nor do they refer to any widely available text as reference for this test. E.g., if you read all of Simon Cozens "Beginning Perl" then you can earn 100% on this test.

Not vendor certified

This Perl test is not certified or ratified by the inventor of Perl, the Perl foundation or any other body related to the governing of Perl.

No ability to flag and comment on questions

Some of the questions were definitely questionable. And some questions had answers which were not entirely right. And most questions were certainly not practical. Now, no feedback on particular questions was possible while taking it. They did solicit feedback at the end, but what good are words without concrete examples to tie them to?

related nodes

certification foo
  • Comment on Some reflections on the Brainbench Perl Test

Replies are listed 'Best First'.
Re: Some reflections on the Brainbench Perl Test
by Your Mother (Chancellor) on Dec 13, 2008 at 01:17 UTC

    Looks like they have a few (3 or more?) Perl tests. This curriculum page makes it look like you got a stripped down version.

    I took an in-house test at a recruiter's office about 1.5 years ago which had two broken questions (concretely wrong answers on things like args to built-ins) and many poor or v4-5.4 practices considered the right answer. I was in shock thinking I'd just lost the job. I complained about a couple of the questions and I was told that quite a few other applicants had said exactly the same thing. I ended up with the job after all. Not a fun experience though.

Re: Some reflections on the Brainbench Perl Test
by mr_mischief (Monsignor) on Dec 13, 2008 at 04:42 UTC
    The Brainbench Perl tests I've taken were a joke, frankly. I'm glad I took them just for sake of curiosity and not because a job depended on them. The certificates look pretty nice.

    I took some other company's Perl test as part of the application for a job a few years ago. I don't remember the name of it, but the questions mostly were much more sane.

    There are, so far as I am aware, no Perl tests endorsed by Larry, The Perl Foundation, or any publisher or author of the better-known books on Perl.

    I'm not sure a certification test meant to be taken in the usual time frame for those sorts of things could even come close to measuring more than glancing Perl knowledge. The language is just too large, complex, and flexible to cover in half an hour or even four hours. All one can be sure of realistically is that a decent test will trip up those who are dishonest about knowing the very basics of the language. Anything past that would require a very good test and perhaps some wishful thinking.

Re: Some reflections on the Brainbench Perl Test
by submersible_toaster (Chaplain) on Dec 13, 2008 at 16:10 UTC

    It's difficult to critique a moving target, as I am not convinced that the Brainbench perl 'quiz' I did in 2007 bears any resemblance to that which you have experienced, and by the sounds of it - there are several perl exams.

    From memory (rusty) it dealt with more of the basic language than idioms or patterns prevalent in perl. I do recall skipping a couple of questions that I regarded as un-answerable per the text. I've also undergone some home-grown perl tests from various employers which generally have a similar focus on depth of understanding of the language and syntax. My favorite question from that pool was "How would you parse a CSV input?" , the answer of course being.. "Don't!, use Text::CSV and get on with your life."

    I can't believe it's not psellchecked
      My favorite question from that pool was "How would you parse a CSV input?" , the answer of course being.. "Don't!, use Text::CSV and get on with your life."

      I've had the same general philosophy overall, but strangely enough that's rarely the answer employers are looking for. So I've come up with a multipart response:

      1. Ask for more details about the CSV being input. Ask the clever things like what types of values could be in the string, what's the encoding, are paired quotes acceptable, is the first row a header, and anything else I think of.
      2. Tell them my first choice would be to use Text::CSV - or if my memory is sharp that day, I'd mention a few of the variants. I'd also try to show them an example of usage appropriate to the information I gathered in part one.
      3. Then, in the inevitable case where they look at me with that "Yes yes, but what if a sniper takes a shot at you every time you load cpan?" look, I'd try to sort through the solution just with core components. It's tedious, and I'd prefer to not take a job where using CPAN was a capital offense. Then again, just looking at it as a problem-solving exercise makes the interview kind of fun. And really, after enough interviews, I'm looking for any way to make them more pleasant experiences.
        "Yes yes, but what if a sniper takes a shot at you every time you load cpan?"

        "I'd have to start looking for another job, sir, as, apparently, this one is far too dangerous to my health."

Re: Some reflections on the Brainbench Perl Test
by converter (Priest) on Dec 13, 2008 at 15:53 UTC
    I don't recall all the details, but a few years ago I had a similar experience with the Brainbench Perl exam and as a result of my list of complaints (with corrections) was invited to serve as a reviewer for a revision of the exam. Several corrections were made to the newly revised exam as a result of my reviews, but Brainbench were not interested in focusing on more advanced topics at the time. I vaguely recall having the impression that Brainbench had determined that customers were only interested in verifying applicants' knowledge of Perl for very basic tasks, and that other languages were more appropriate for real application development. Unfortunate.
Re: Some reflections on the Brainbench Perl Test
by CountZero (Bishop) on Dec 13, 2008 at 22:46 UTC
    I think I said it before, Perlmonks should issue Perl-certificates, not based upon a silly quiz, but based upon the level of Perl knowledge shown in answering and posing questions. I'm sure one of our Monks can whip-up a module to print nice certificates to be signed by the "Chancellor of the Certification Division" or the "Curate of the Inquisitory Board into Practical Perl knowledge".


    A program should be light and agile, its subroutines connected like a string of pearls. The spirit and intent of the program should be retained throughout. There should be neither too little or too much, neither needless loops nor useless variables, neither lack of structure nor overwhelming rigidity." - The Tao of Programming, 4.1 - Geoffrey James

      I can think of a pretty basic algorithm to make this fly. First, you probably discard levels/experience as showing up to vote can place you among the mighty given enough time. Then the ratings of any post with lines of code are measured positive against negative (not sure how or in what formula). A level of "competent" is determined by identifying a monk who is recognized as such, but not by much :), and then peg the scale to that code:positive/negative node rating measure. Anyone above is PM Certifiedô. Categories of certification can be tag mapped from the use statements in a monk's posted code samples. E.g.: considering my own posts, I would expect to be auto-certified for CGI/Ajax but not Sysadmin.

      All y'all are lucky I'm too busy to try to make this into working code right now.

Re: Some reflections on the Brainbench Perl Test
by chromatic (Archbishop) on Dec 13, 2008 at 23:39 UTC

    If I were hiring programmers, I'd give them the Assess Your Agility quiz before checking language certifications.

        Agility is a myth.

        In that case, how can anyone argue with such a well-defended and diligently researched assertion as that?

        Tuesday will be the 24th stable monthly release of Parrot in a row. We don't score 100% on that quiz, but I'm happy to compare our score now versus three years ago and put that up against your "evidence" any day.

        I don't pretend to disagree but after reading the evaluations questions at the link provided, I would *much* rather work somewhere that could score well on the assessment than not.

        I want to know if someone is willing to work on a team like that. The interviewee gets extra credit for wanting to work on a team like that.

Re: Some reflections on the Brainbench Perl Test
by sundialsvc4 (Abbot) on Dec 16, 2008 at 00:59 UTC

    I respect BrainBench, in a caveat emptor sort of way, for being as successful as they have been at dishing-out pure crap :-) and somehow making money at it.

Log In?

What's my password?
Create A New User
Node Status?
node history
Node Type: perlmeditation [id://730066]
Approved by kyle
[Discipulus]: Eternal City?

How do I use this? | Other CB clients
Other Users?
Others pondering the Monastery: (7)
As of 2017-02-28 10:30 GMT
Find Nodes?
    Voting Booth?
    Before electricity was invented, what was the Electric Eel called?

    Results (399 votes). Check out past polls.