The next generation and cheating

Discussion in 'MCSD' started by Justin Weinberg, Feb 6, 2006.

  1. I'm curious what MS is actively doing to curtail cheating with the next
    generation of certs?

    My suggestions are:

    1. A minimum question bank of one thousand questions per exam, adding an
    additional 100 questions every two months. These don't have to be killer
    difficult questions. In fact I would divide questions into two groups. The
    first group would be "gimme" questions of which this thousand would belong.
    This makes it brain dead easy to keep the size of the question bank high.
    Then have a secondary set of questions of which you maybe have 100 which are
    tough questions, rotating as often as possible. Each exam should be 70%
    gimmes and 30% toughies. Scores would be weighted as a percentage of gimmes
    to toughies.

    If you do this, you remove the cheating incentive, because it would be
    easier to get an honest understanding of the topics than to memorize 1000
    questions.

    2. Remove the cheating incentive by minimizing the amount of esoteric stuff
    in the tests. I failed the SQL exam the first time, but even after
    extensive studying, I only made a 700. Does this mean I do not have a high
    experience level with SQL Server? I think not. I think it means that
    whether I can come up with the correct DBCC command without referencing
    books online is meaningless.

    3. Monitor braindump sites extensively. Flood them with incorrect
    questions and answers. Make the information unreliable.

    4. Make people buy into honesty. As part of the testing process, make
    people sign a secondary agreement of honesty or take a verbal oath - or
    anything to get them to treat the tests with respect.

    5. Use technology - we are technologists, right? Craft dynamic questions,
    where the computer can randomly pick from 10 or so options to inject in a
    problem, and have an algorithm that translates options to correct answers.
    This in itself would increase the complexity of the exam by a factorial.

    I have one test left for my MCSD, and a thousand dollars and hundreds of
    hours later, I question the value of it because I keep running into people
    with certs that have no business having them.

    Do this generation of certs right MS.
    Justin Weinberg, Feb 6, 2006
    #1
    1. Advertising

  2. Justin Weinberg

    JaR Guest

    Echoed forth from the dank caverns of microsoft.public.cert.exam.mcsd,
    the plaintive wail of Justin Weinberg:

    > I'm curious what MS is actively doing to curtail cheating with the next
    > generation of certs?


    Fsck-all.

    <rant at the wind snipped>

    Sux, don't it? They could not care less. It's nothing more than a money
    farm for Bill, and a bloody genius bit of marketing. If they discouraged
    cheats, they'd sell less tests. We have reported cheats to MS, we have
    notified them of cheat sites, I have even gone so far as to e-mail and
    ask whether they condoned or condemned certain cheat sites.

    Do you know what response I (or anybody I know) has gotten from
    Microsoft?

    Care to guess?

    Anyone? Anyone? Class? Anyone?

    Nada, Zip, Zero, in short; fsck-all.

    --
    JaR
    MCNGP 10110
    Remove hat to reply
    The only worthwhile doors in life that will open for you are the ones
    you kick down or make for yourself. ~Rob Lind man
    JaR, Feb 8, 2006
    #2
    1. Advertising

  3. Hello.

    Try brainbench.com
    .... they have a better solution: 40 out of about 100-150 questions (I
    guess), differentiated difficulty on the questions, the better you answer,
    the harder the questions get, questions are invisibly classified something
    like "easy", "not easy" and "hard", where "hard" questions often are
    confusing or based on niche-knowledge (much like this post); ca. 50% correct
    is enough to pass, ca. 75% gives you something called the master-level. The
    system is not perfect, some of the tests are so hard that none has made the
    master-level and others so easy that too many can get the master-level. I
    think it is expected that about 40% shall pass, 20-25% of these with master
    level.

    Suggestion
    1) Questions should be updated regulary, without notice.
    2) You could require at least one master-level pass to get any title.
    3) There might be something called Master Professional with additional
    benefits (there are fewer and fewer jobs in the western world and such ways
    to differentiate on self could be a selling-point for the specialists)
    4) You should not get your score reported, only "failed", "passed" or
    "master" (with the possible addition of "don't go anywere; you need to come
    work for us")
    Braindump-business should need to take more tests to get all the knowledge
    and not as much basis on how much of a test material they understand.
    It would be challenging to create braindumps helpful to pass any exam at
    master level, meaning that even if people used braindumps, they would need
    to master at least one area to get a title (see 2).

    Expected results
    This would, in my opinion, not only kill the braindump-business, but also
    help to give additional (automatic) general advice to test-takers once they
    get their, say MCSD, for instance:
    - "BizTalk speciallity adviced" (because of strong integration and process
    skills)
    - "Go for Certified Architect" (because of strong
    business/scenarion-understanding)
    - "Add DBA-skill to you resume" (because of strong database-skills)
    - "Add security/network-skill to you resume" (because of what ever is needed
    here)
    - "..." (because of ...)

    Cost
    This would cost money. Money must come from somewhere. Preferably earnings.
    The additional cost of this way of doing things could be covered by
    selling/feeding information to reqruiters and sharing prospects with
    internal MS and partners.
    Helping in the process of matching people against job requirements.
    Also, if it could help msdn and technet to develop documentation on areas no
    one understands (say XQvery for instance, or maybe I'm the only one that
    can't even spell it...), it might be seen as an investment in the
    documentation.

    Gorm Braarvig

    "Justin Weinberg" <jweinberg@_spammyoff_mrgsoft.com> wrote in message
    news:...
    > I'm curious what MS is actively doing to curtail cheating with the next
    > generation of certs?
    >
    > My suggestions are:
    >
    > 1. A minimum question bank of one thousand questions per exam, adding an
    > additional 100 questions every two months. These don't have to be killer
    > difficult questions. In fact I would divide questions into two groups.
    > The first group would be "gimme" questions of which this thousand would
    > belong. This makes it brain dead easy to keep the size of the question
    > bank high. Then have a secondary set of questions of which you maybe have
    > 100 which are tough questions, rotating as often as possible. Each exam
    > should be 70% gimmes and 30% toughies. Scores would be weighted as a
    > percentage of gimmes to toughies.
    >
    > If you do this, you remove the cheating incentive, because it would be
    > easier to get an honest understanding of the topics than to memorize 1000
    > questions.
    >
    > 2. Remove the cheating incentive by minimizing the amount of esoteric
    > stuff in the tests. I failed the SQL exam the first time, but even after
    > extensive studying, I only made a 700. Does this mean I do not have a
    > high experience level with SQL Server? I think not. I think it means
    > that whether I can come up with the correct DBCC command without
    > referencing books online is meaningless.
    >
    > 3. Monitor braindump sites extensively. Flood them with incorrect
    > questions and answers. Make the information unreliable.
    >
    > 4. Make people buy into honesty. As part of the testing process, make
    > people sign a secondary agreement of honesty or take a verbal oath - or
    > anything to get them to treat the tests with respect.
    >
    > 5. Use technology - we are technologists, right? Craft dynamic
    > questions, where the computer can randomly pick from 10 or so options to
    > inject in a problem, and have an algorithm that translates options to
    > correct answers. This in itself would increase the complexity of the exam
    > by a factorial.
    >
    > I have one test left for my MCSD, and a thousand dollars and hundreds of
    > hours later, I question the value of it because I keep running into people
    > with certs that have no business having them.
    >
    > Do this generation of certs right MS.
    >
    Gorm Braarvig, Feb 9, 2006
    #3
  4. Justin Weinberg

    Frisbee® Guest

    "Gorm Braarvig" <> wrote in message
    news:uH%23hs%...
    > Hello.
    >
    > Try brainbench.com


    I don't know whether to laugh or cry.
    Frisbee®, Feb 9, 2006
    #4
  5. Justin Weinberg

    JaR Guest

    Echoed forth from the dank caverns of microsoft.public.cert.exam.mcsd,
    the plaintive wail of Gorm Braarvig:

    > Helping in the process of matching people against job requirements.
    > Also, if it could help msdn and technet to develop documentation on
    > areas no one understands (say XQvery for instance, or maybe I'm the
    > only one that can't even spell it...), it might be seen as an
    > investment in the documentation.


    In the immortal words of an old friend:

    "You just don't understand! It doesn't scaaaaaale!"~$.50 to L.A.R.

    --
    JaR
    MCNGP 10110
    Remove hat to reply
    Have you noticed that a slight tax increase costs you two hundred dollars,
    and a substantial tax cut saves you thirty cents?
    JaR, Feb 9, 2006
    #5
  6. >> Helping in the process of matching people against job requirements.
    >> Also, if it could help msdn and technet to develop documentation on
    >> areas no one understands (say XQvery for instance, or maybe I'm the
    >> only one that can't even spell it...), it might be seen as an
    >> investment in the documentation.

    >
    > In the immortal words of an old friend:
    >
    > "You just don't understand! It doesn't scaaaaaale!"~$.50 to L.A.R.


    What does not scale?

    - XQuery
    Agreed, but XQuery against a field of a record of a recordset in SQL2005
    does not necessary need to scale.

    - Matching people against job requirements
    No. Does not scale. There is room for smart solutions here. Maybe you are
    the guy to invent something?

    - develop documentation on areas no one understands
    Scales?

    BTW: do this scale: http://62.89.36.69/g/g.html (Suspicios site)?
    Scalability is a special interest of mine, if you find something that does
    not scale, it might be something I can fix.

    > You just don't understand!

    You are right. I might be a bit slow. Please explain, I am into learning.
    Gorm Braarvig, Feb 9, 2006
    #6
  7. "Frisbee®" <> wrote in message
    news:%...
    > "Gorm Braarvig" <> wrote in message
    > news:uH%23hs%...
    >> Hello.
    >>
    >> Try brainbench.com

    >
    > I don't know whether to laugh or cry.
    >
    >


    Knowledge is over-rated.
    Gorm Braarvig, Feb 9, 2006
    #7
  8. Justin Weinberg

    JaR Guest

    Echoed forth from the dank caverns of microsoft.public.cert.exam.mcsd,
    the plaintive wail of Gorm Braarvig:

    >> You just don't understand!

    > You are right. I might be a bit slow. Please explain, I am into
    > learning.


    It's a quote from a discussion we had long ago with one who allegedly had
    inside info on the exam prep process. Her constant reply to various
    suggestions about the testing procedure was always that we did not
    understand the arcane machinations of the test architects, and that our
    suggestions were invalid, as they just would not scale.

    The one to explain has not been around for a while, and probably
    fortunately, as she would be just as unable to explain why $Solution does
    not scale, but would be repeatedly telling us all what dolts we are for
    not understanding in the first place.

    I leave it to others more masochistic than I to invoke L.A.R. if they
    feel the need for further clarification.

    --
    JaR
    MCNGP 10110
    Remove hat to reply
    If you're playing a poker game and you look around the table and can't
    tell who the sucker is, it's you.
    JaR, Feb 9, 2006
    #8
  9. "JaR" <> wrote in message
    news:Xns9765744022213Talleywhacker@207.46.248.16...
    > Echoed forth from the dank caverns of microsoft.public.cert.exam.mcsd,
    > the plaintive wail of Gorm Braarvig:
    >
    >>> You just don't understand!

    >> You are right. I might be a bit slow. Please explain, I am into
    >> learning.

    >
    > It's a quote from a discussion we had long ago with one who allegedly had
    > inside info on the exam prep process. Her constant reply to various
    > suggestions about the testing procedure was always that we did not
    > understand the arcane machinations of the test architects, and that our
    > suggestions were invalid, as they just would not scale.
    >
    > The one to explain has not been around for a while, and probably
    > fortunately, as she would be just as unable to explain why $Solution does
    > not scale, but would be repeatedly telling us all what dolts we are for
    > not understanding in the first place.
    >
    > I leave it to others more masochistic than I to invoke L.A.R. if they
    > feel the need for further clarification.
    >
    > --
    > JaR
    > MCNGP 10110
    > Remove hat to reply
    > If you're playing a poker game and you look around the table and can't
    > tell who the sucker is, it's you.


    I understand.

    Most people I have met with the "will not scale"-argument
    a) does not know what they are talking about
    b) have no low-level experience
    c) have no math-background
    d) have low problem-solving skills
    e) talks mostly in general terms
    f) has abstract reasons ready for anything and specific knowledge of nothing
    ....the list goes on...

    the point I want to make is that ANYTHING is scalable. Many things are hard
    to scale, "arcane machinations of" quote on quote on quote on quote "test
    architects" can't possibly qualify to be hard to scale. This discussion
    would leave an impression on me too, just sad I could not be part of it,
    sounds out of this world...

    BTW:
    Only exp I have on these tests are on betas on the new TS and PRO-exams, and
    I don't think they are too bad (the questions, the composition I know
    nothing of). I quite liked the scenarios with 5-10 questions bound to real
    world scenarios (on some DB-exams). I might change opinion when I get the
    results, of course (I am a soar looser).
    Gorm Braarvig, Feb 9, 2006
    #9
  10. Justin Weinberg

    Frisbee® Guest

    "JaR" <> wrote in message
    news:Xns9765744022213Talleywhacker@207.46.248.16...
    > Echoed forth from the dank caverns of microsoft.public.cert.exam.mcsd,
    > the plaintive wail of Gorm Braarvig:
    >
    >>> You just don't understand!

    >> You are right. I might be a bit slow. Please explain, I am into
    >> learning.

    >
    > It's a quote from a discussion we had long ago with one who allegedly had
    > inside info on the exam prep process. Her constant reply to various
    > suggestions about the testing procedure was always that we did not
    > understand the arcane machinations of the test architects, and that our
    > suggestions were invalid, as they just would not scale.
    >
    > The one to explain has not been around for a while, and probably
    > fortunately, as she would be just as unable to explain why $Solution does
    > not scale, but would be repeatedly telling us all what dolts we are for
    > not understanding in the first place.
    >
    > I leave it to others more masochistic than I to invoke L.A.R. if they
    > feel the need for further clarification.


    I think she might have just been on the rag or something.

    I never understood it either.

    But I was askeered to ask her anymore.
    Frisbee®, Feb 10, 2006
    #10
  11. Justin Weinberg

    Frisbee® Guest

    "Gorm Braarvig" <> wrote in message
    news:...
    > "JaR" <> wrote in message
    > news:Xns9765744022213Talleywhacker@207.46.248.16...
    >> Echoed forth from the dank caverns of microsoft.public.cert.exam.mcsd,
    >> the plaintive wail of Gorm Braarvig:
    >>
    >>>> You just don't understand!
    >>> You are right. I might be a bit slow. Please explain, I am into
    >>> learning.

    >>
    >> It's a quote from a discussion we had long ago with one who allegedly had
    >> inside info on the exam prep process. Her constant reply to various
    >> suggestions about the testing procedure was always that we did not
    >> understand the arcane machinations of the test architects, and that our
    >> suggestions were invalid, as they just would not scale.
    >>
    >> The one to explain has not been around for a while, and probably
    >> fortunately, as she would be just as unable to explain why $Solution does
    >> not scale, but would be repeatedly telling us all what dolts we are for
    >> not understanding in the first place.
    >>
    >> I leave it to others more masochistic than I to invoke L.A.R. if they
    >> feel the need for further clarification.
    >>
    >> --
    >> JaR
    >> MCNGP 10110
    >> Remove hat to reply
    >> If you're playing a poker game and you look around the table and can't
    >> tell who the sucker is, it's you.

    >
    > I understand.
    >
    > Most people I have met with the "will not scale"-argument
    > a) does not know what they are talking about
    > b) have no low-level experience
    > c) have no math-background
    > d) have low problem-solving skills
    > e) talks mostly in general terms
    > f) has abstract reasons ready for anything and specific knowledge of
    > nothing
    > ...the list goes on...


    None of those apply to the person in question. She knew her sh!t, there was
    never any doubt about that. We were stupid and that was that (seriously).
    Frisbee®, Feb 10, 2006
    #11
  12. Justin Weinberg

    JaR Guest

    Echoed forth from the dank caverns of microsoft.public.cert.exam.mcsd, the
    plaintive wail of Frisbee®:

    > I think she might have just been on the rag or something.
    >
    > I never understood it either.
    >
    > But I was askeered to ask her anymore.


    Pfft. Having a discussion with L. on the best of days was like french
    kissing a land mine.

    Although it was fun watching the odd goofball trying to outlast her at
    getting in the "last word". Usta remind me of my grandkids doing the same
    thing. LOL!

    --
    JaR
    MCNGP 10110
    Remove hat to reply
    Anyone who puts a small gloss on a fundamental technology, calls it
    proprietary, and then tries to keep others from building on it, is a thief.
    ~Tim O'Reilly
    JaR, Feb 10, 2006
    #12
  13. >>> If you're playing a poker game and you look around the table and can't
    >>> tell who the sucker is, it's you.


    he, he, I finally got it, the joke is on me.

    > None of those apply to the person in question. She knew her sh!t, there
    > was never any doubt about that. We were stupid and that was that
    > (seriously).


    let's hope she returns, I have always liked chating with the god-like people
    that doesen't need to bow for logic.
    Gorm Braarvig, Feb 11, 2006
    #13
  14. Justin Weinberg

    JaR Guest

    Ping: Laura A. Robinson was: Re: The next generation

    Echoed forth from the dank caverns of microsoft.public.cert.exam.mcsd,
    the plaintive wail of Gorm Braarvig:

    > let's hope she returns, I have always liked chating with the god-like
    > people that doesen't need to bow for logic.


    Be careful what you ask for.

    --
    JaR
    MCNGP 10110
    Remove hat to reply
    How would a car function if it were designed like a computer?
    Occasionally, executing a maneuver would cause your car to stop and fail
    and you would have to re-install the engine, and the airbag system would
    say, "Are you sure?" before going off. ~Katie Hafner
    JaR, Feb 13, 2006
    #14
  15. Justin Weinberg

    Kline Sphere Guest

    >Her constant reply to various
    >suggestions about the testing procedure was always that we did not
    >understand the arcane machinations of the test architects, and that our
    >suggestions were invalid, as they just would not scale.


    Well that's true, I don't 'understand the arcane machinations of the
    test architects'. My point at the time, and I've not changed my view
    since, is that the [current] process is crap. The process 'certifies'
    nothing because the certification format is open to abuse because
    people can and do cheat. Other organizations, all with less financial
    clout the microsoft manage the provide a certification process that
    works, i.e. a process that involves a written and/or lab based core
    part of the process, whereby the candidate has to explain and describe
    not only the answer, but why they came about the answer. Unlike ms
    software, this process will scale :)

    Kline Sphere (Chalk) MCNGP #3
    Kline Sphere, Feb 15, 2006
    #15
  16. Justin Weinberg

    JaR Guest

    Echoed forth from the dank caverns of microsoft.public.cert.exam.mcsd, the
    plaintive wail of Kline Sphere:

    > Other organizations, all with less financial
    > clout the microsoft manage the provide a certification process that
    > works, i.e. a process that involves a written and/or lab based core
    > part of the process, whereby the candidate has to explain and describe
    > not only the answer, but why they came about the answer. Unlike ms
    > software, this process will scale :)


    That's what made L's constant protestations of lack of scalability so
    aggravating.

    No other certification body that I have ever had contact with seems to care
    so little about the validity of their certifying process.

    Plumbers have a far tougher examination of their skills.

    --
    JaR
    MCNGP 10110
    Remove hat to reply
    The plural of anecdote is not data. ~Frank Kotsonis
    JaR, Feb 15, 2006
    #16
  17. "JaR" <> wrote in message
    news:Xns976B921B178B8Talleywhacker@207.46.248.16...
    > Echoed forth from the dank caverns of microsoft.public.cert.exam.mcsd, the
    > plaintive wail of Kline Sphere:
    >
    >> Other organizations, all with less financial
    >> clout the microsoft manage the provide a certification process that
    >> works, i.e. a process that involves a written and/or lab based core
    >> part of the process, whereby the candidate has to explain and describe
    >> not only the answer, but why they came about the answer. Unlike ms
    >> software, this process will scale :)

    >
    > That's what made L's constant protestations of lack of scalability so
    > aggravating.
    >
    > No other certification body that I have ever had contact with seems to
    > care
    > so little about the validity of their certifying process.
    >
    > Plumbers have a far tougher examination of their skills.
    >
    > --
    > JaR
    > MCNGP 10110
    > Remove hat to reply
    > The plural of anecdote is not data. ~Frank Kotsonis


    I disagree. My limited exp with the exams tells me they are hard.

    I've only got the result on one beta exam (database I believe), I failed.
    I think I am pretty good at databases; I also have two years of exp on SQL
    2005, not enough.
    I had no test-specific preps, but I thought that, being the guy that I am, I
    might pass anyway.
    The thought of some inexperienced moron memorizing the answers and pass this
    exam makes me think I wasted most of the time taking it (a full day because
    of problems with the exam).
    Tests should be dynamic by nature.
    As an occasional recruiter, I would (even) prefer brainbench on-site to
    looking for MC*s. Can't someone just invent something?
    Gorm Braarvig, Feb 16, 2006
    #17
  18. > I disagree. My limited exp with the exams tells me they are hard.

    Sorry, JaR, and greetings!
    You were talking about the [current] exams.
    I don't disagree, I know nothing.
    Gorm Braarvig, Feb 16, 2006
    #18
  19. Justin Weinberg

    JaR Guest

    Echoed forth from the dank caverns of microsoft.public.cert.exam.mcsd,
    the plaintive wail of Gorm Braarvig:

    >
    >> I disagree. My limited exp with the exams tells me they are hard.

    >
    > Sorry, JaR, and greetings!
    > You were talking about the [current] exams.
    > I don't disagree, I know nothing.


    Well, to be fair, it's not the exam itself to which I was referring. My
    comment about plumbers is meant to reference the fact that they are
    required to serve a period of apprenticeship and/or technical school
    training before they are even _allowed_ to take any certification exam.
    And that exam has a practical, hands-on component.

    Any fsckwit with the correct fees paid may take any Microsoft exam and
    become "certified" as an expert in a field that in all likelihood, he may
    never have worked a single day.

    --
    JaR
    MCNGP 10110
    Remove hat to reply
    Every decent man is ashamed of the government he lives under. ~H. L.
    Mencken
    JaR, Feb 17, 2006
    #19
  20. Justin Weinberg

    Lydon Bergin Guest

    > Any fsckwit with the correct fees paid may take any Microsoft exam and
    > become "certified" as an expert in a field that in all likelihood, he may
    > never have worked a single day.


    While that is true, it would also be a little overzealous, not to mention a
    logistical nightmare, to only allow people who have worked in the field to
    take certification exams. I am working on my MCAD Certification and I have
    passed one of the exams and NEVER WORKED AS A DEVELOPER, but that's what I
    eventually want to do. That sure doesn't mean that I don't know how to
    code, because I do; it just means that because you have to have experience
    to get a job, and you have to have a job to get experience, I haven't been
    able to get a job! But, since I've gotten my MCP about a month ago, I have
    already gotten an interview for a development job and I knew everything that
    the interviewer asked me.

    If MS decides to make the next generation only available to people that have
    worked in the field, then that will make the Cache 22 of trying to get a job
    even worse. Sure, there are people who just memorize BrainDumps, but there
    are also those of us that take the time to actually learn the objectives to
    pass the exams, and we should be able to do that.
    Lydon Bergin, Feb 17, 2006
    #20
    1. Advertising

Want to reply to this thread or ask your own question?

It takes just 2 minutes to sign up (and it's free!). Just click the sign up button to choose a username and then you can ask your own questions on the forum.
Similar Threads
  1. Silverstrand
    Replies:
    0
    Views:
    642
    Silverstrand
    Jan 10, 2006
  2. Replies:
    0
    Views:
    391
  3. Sunil Sood

    BT's Next Generation Consumer Strategy

    Sunil Sood, Dec 27, 2005, in forum: UK VOIP
    Replies:
    0
    Views:
    488
    Sunil Sood
    Dec 27, 2005
  4. John123

    What to expect from next generation Canon IXUS?

    John123, Nov 16, 2003, in forum: Digital Photography
    Replies:
    0
    Views:
    289
    John123
    Nov 16, 2003
  5. Allan
    Replies:
    0
    Views:
    6,005
    Allan
    Jul 15, 2005
Loading...

Share This Page