April 4, 2010 at 6:04 am
Chris-232075 (3/23/2010)
I'm generally pro-certification, but specifically anti-Microsoft certification since so much of the information learned for passing their exams are product specific, and therefore, obsolete in relatively short order.
It's not just MS certification that's bad, it's pretty well all industry certification. It will remain bad as long as training companies can pretty well guarantee to get even the dimmest bulb though th eexam on two tries, so that they can offer an "if you fail we'll train you again, if you fail three times you get your money back" deal and make large profits on it; and of course this increases the volume of examination candidates, and hence the examination fee income for the providers of the certifications, so they have a perverse incentive to make the exams easy to pass. I think MS has begun to try to climb out of this pit (but is doing so very slowly indeed, so it's not clear that they really intend to get rid of the problem) but most of the others haven't.
The IEEE Computer Society certifications are examples of excellent certifications that aren't software specific and that I'd like to see American professionals explore more fully.
I followed the link and was quite unpleasantly surprised.
CSDA is an exam for either people in their final undergrad year or people with no higher education and two years of programming, and isn't going to tell us anything about anything other than the ability to study the CSDA e-learning "online, self-paced" course (4 modules, 500 practise questions) and pass the exam after doing that; and of course the ability to find $645 plus tax to pay for it. In what way should that be taken more seriously than MCSE, for example? I can't see it as anything other than a money-spinner for IEEE.
CSDP is aimed at two years experience after CSDA for the non-graduates or 2 years after masters degree of 4 years after bachelors degree. That looks to me like age early to mid twenties, but IEEE call it a mid-career qualification (if my mid-carrer had been at that age I'd have retired about 40 years ago).
Neither certification appears to require any form of reference from qualified persons - they are purely examination based. That's outrageous for a qualification offered by a professional society.
Tom
April 5, 2010 at 9:27 am
"CSDA is an exam for either people in their final undergrad year or people with no higher education and two years of programming, and isn't going to tell us anything about anything other than the ability to study the CSDA e-learning "online, self-paced" course (4 modules, 500 practise questions) and pass the exam after doing that; and of course the ability to find $645 plus tax to pay for it. In what way should that be taken more seriously than MCSE, for example?"
Answering this question, the difference is content? One is product specific, and the other is not.
I don't see how any exam can be structured that doesn't tell us about one's ability to "study the course" and "pass the exam"--unless we prefer a certification with an unusually small pass rate, and therefore, an unusually small number of people who even bother to take it. Taking Tom's logic to extremes, we can say the same thing about nearly all undergraduate and many graduate degrees (i.e., such degrees only tell us about a person's ability to pay the money, take a course, and pass a test). In fact, learning the content is beneficial. I support certifications that require experience, but I don't dismiss those that don't.
Although the IEEE exams are not database exams, I offered them up as examples of certifications that provide content covering fundamental material, NOT product specific, that many either never learned (I, for one, have a degree in business--not IT) or learned and forgot. Generally speaking, the material does not become obsolete over time.
(note: Because of technical difficulties with this site, I had to get a new login. I am the commenter formerly known as Chris.)
April 5, 2010 at 10:04 am
"Neither [IEEE] certification appears to require any form of reference from qualified persons - they are purely examination based. That's outrageous for a qualification offered by a professional society."
I have to disagree with Tom on this one as well. Beyond telling us years of experience in a certain position, references, as far as I can tell, can be abused or an impossible hindrance since the certification would rely on managers who might lack the required maturity, objectivity, or knowledge (e.g., my managers do not have IT training or work experience).
With regard to experience, my wife is in a field where, in order to be licensed, they must track and maintain experience in specific areas of their field. Through the years, most (not all) of her co-workers plug in bogus numbers only to have managers sign off on them without really caring whether or not the figures are accurate (it seems to be the norm). Others have managers who won't sign off on false data, but then the managers won't give them the experience in needed areas because they don't care about helping employees achieve the licensing requirements.
Yes, I've been a software developer/database developer for ten years. My employers can verify that. Beyond that, we'd be asking for trouble. We need to certify a person's knowledge and skills--not whether or not subjective managers will come through for an ambitious employee who wants to be certified.
Fortunately, most industries have few, if any, certifications requiring references. As far as I'm concerned, that's the way it should be.
April 5, 2010 at 10:16 am
I think that any exam can be gamed to an extent, and despite all the efforts put into doctors, we still turn out poor doctors because the real world isn't an exam.
However we can do better, and an exam is something we can design better. I'm not sure I like the essay portion of the exam, since some people cannot explain themselves well in writing. I guess some don't do well in interviews either. Maybe we have an interview portion that's oral or written?
In terms of a cert, I think we do the best we can, which is get someone to attest to their experience, jobs, education, etc. Someone will game it, but when they get caught, it's on them, not the certifers.
Follow me on Twitter: http://www.twitter.com/way0utwest
Forum Etiquette: How to post data/code on a forum to get the best help
My Blog: www.voiceofthedba.com
April 5, 2010 at 10:25 am
Steve Jones - Editor (4/5/2010)
I think that any exam can be gamed to an extent, and despite all the efforts put into doctors, we still turn out poor doctors because the real world isn't an exam.However we can do better, and an exam is something we can design better. I'm not sure I like the essay portion of the exam, since some people cannot explain themselves well in writing. I guess some don't do well in interviews either. Maybe we have an interview portion that's oral or written?
In terms of a cert, I think we do the best we can, which is get someone to attest to their experience, jobs, education, etc. Someone will game it, but when they get caught, it's on them, not the certifers.
Building on a thought of Steve's here (at least he provoked the thought).
No exam can cover absolutely every scenario that could be encountered in the real world. I don't believe it to be possible, and if it is possible - it is not feasible.
I don't like requiring somebody to sit for 4+ hours of examination at a single time, and then to require it for each exam. That throws the cost too high for this middle ground cert, and also creates a prohibitive barrier. This time requirement for examination becomes too much like the MCM if implemented to this extreme.
I also don't think that an essay would serve appropriately here. The only purpose of an essay, IMHO, is to show communications skills. This also requires a lot more time to score the exam. For the communications skills assessment, we are already looking at implementing a presentation and a review panel that should include an interview with the candidate.
Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events
April 5, 2010 at 12:51 pm
We have covered TIS. I think that can be tabled.
How about we move on to a new topic?
I think the next viable topic would be to hash out how the review board should be handled.
So, how will we propose this particular requirement?
Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events
April 5, 2010 at 1:51 pm
CirquedeSQLeil (4/5/2010)
We have covered TIS. I think that can be tabled
I just have one concern about the TIS: the vote was asked for on Good Friday, and there haven't been a lot of votes, especially from people that have posted thoughts about it. I'm concerned that the two are related... I think we may need to keep it open for another day or two, just to get their input.
Wayne
Microsoft Certified Master: SQL Server 2008
Author - SQL Server T-SQL Recipes
April 5, 2010 at 1:54 pm
Has we handle the different topics, are we going to have a central place to have this all written down? Jason, are you going to keep your blog up-to-date as we finish a topic, and if so is this an option for having a central place for the results?
Wayne
Microsoft Certified Master: SQL Server 2008
Author - SQL Server T-SQL Recipes
April 5, 2010 at 2:24 pm
GilaMonster (3/31/2010)
Lynn Pettis (3/31/2010)
GilaMonster (3/31/2010)
WayneS (3/31/2010)
dma-669038 (3/31/2010)
One guy i talked to did not know the difference between table scan and index seek except that 'it ran faster')Makes me wonder what he thinks the difference is between a table scan and a clustered index scan...
That's an evil question to ask.... So many ways to get it wrong.
Tell me if I'm wrong, but I've always thought that they were the same, except you'd get a table scan on a heap and a clustered index scan where a clustered index existed.
Exactly.
Problem is, too many people think table scan = slow and clustered index scan, because it says index, is something much better and preferable, and don't realise they are essentially the same thing.
One I've heard before -
Q: "Which is better, a table scan or a clustered index scan?"
A: "The clustered index scan"
Q: "Why?"
A: "Because it's using the index and the table scan isn't."
Can I just answer all of these interviews questions with "It Depends ..."?
😀
[font="Times New Roman"]-- RBarryYoung[/font], [font="Times New Roman"] (302)375-0451[/font] blog: MovingSQL.com, Twitter: @RBarryYoung[font="Arial Black"]
Proactive Performance Solutions, Inc. [/font][font="Verdana"] "Performance is our middle name."[/font]
April 5, 2010 at 2:25 pm
WayneS (4/5/2010)
Has we handle the different topics, are we going to have a central place to have this all written down? Jason, are you going to keep your blog up-to-date as we finish a topic, and if so is this an option for having a central place for the results?
I will update the blog. As we conclude a topic, I will round up the information and post it there. As for the prior topic, Voting is till open (so is discussion if people really want to discuss it, but the discussion waned significantly so I thought it proper to start with the next item of business).
Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events
April 5, 2010 at 2:25 pm
RBarryYoung (4/5/2010)
GilaMonster (3/31/2010)
Lynn Pettis (3/31/2010)
GilaMonster (3/31/2010)
WayneS (3/31/2010)
dma-669038 (3/31/2010)
One guy i talked to did not know the difference between table scan and index seek except that 'it ran faster')Makes me wonder what he thinks the difference is between a table scan and a clustered index scan...
That's an evil question to ask.... So many ways to get it wrong.
Tell me if I'm wrong, but I've always thought that they were the same, except you'd get a table scan on a heap and a clustered index scan where a clustered index existed.
Exactly.
Problem is, too many people think table scan = slow and clustered index scan, because it says index, is something much better and preferable, and don't realise they are essentially the same thing.
One I've heard before -
Q: "Which is better, a table scan or a clustered index scan?"
A: "The clustered index scan"
Q: "Why?"
A: "Because it's using the index and the table scan isn't."
Can I just answer all of these interviews questions with "It Depends ..."?
😀
Well, that just depends.
Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events
April 5, 2010 at 2:29 pm
I discussed this with some of my UG members informally. Majority feel 1.5 years of TIS is very less and it is easier to do away with it instead. Some people pointed out that the FAQ for MCM states the average person trying the MCM has over 10 years of SQL experience, so a mid level cert should atleast be half of that (5 years). Again this goes back to the can of worms on how do you know someone has 5 years and people gaming the thing and so on. Just passing that on for what it is worth.
April 5, 2010 at 2:33 pm
Paul White NZ (4/1/2010)
GSquared (4/1/2010)
Could also happen if the table is in a read-only file or datbase, which has the same effect as read uncommitted but without the risks associated with concurrent page splits, et al.And if there are at least 64 pages to read 🙂
I love this question!
Couldn't you just ask "Is your name Paul Randal?" 😛
[font="Times New Roman"]-- RBarryYoung[/font], [font="Times New Roman"] (302)375-0451[/font] blog: MovingSQL.com, Twitter: @RBarryYoung[font="Arial Black"]
Proactive Performance Solutions, Inc. [/font][font="Verdana"] "Performance is our middle name."[/font]
April 5, 2010 at 2:38 pm
dma-669038 (4/5/2010)
I discussed this with some of my UG members informally. Majority feel 1.5 years of TIS is very less and it is easier to do away with it instead. Some people pointed out that the FAQ for MCM states the average person trying the MCM has over 10 years of SQL experience, so a mid level cert should atleast be half of that (5 years). Again this goes back to the can of worms on how do you know someone has 5 years and people gaming the thing and so on. Just passing that on for what it is worth.
I like what you have done with this proposal. Taking this to the UG and discussing it with even more people - very good.
What if we said that the requirement were 1.5 yrs but the average applicant had 5+ yrs experience? Even though the MCM has a much higher time in service average, the requirement is maintained at 5yrs.
Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events
April 5, 2010 at 2:43 pm
Another piece of the puzzle that I just learned today is that each applicant does have to undergo an interview when applying for the MCM. I knew there was a formal application and a review, did not know there was an interview as well. To paraphrase Brent Ozar[/url], one might be able to cheat the interview - but the remaining lab would flush that out.
The exams, application, interview, and lab are rigorous enough that a gamer just would not have a very good chance of achieving an MCM. I envision that the same sort of thing could be said of the MCJ - just not quite as intense as the Master.
Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events
Viewing 15 posts - 631 through 645 (of 685 total)
You must be logged in to reply to this topic. Login to reply