April 2, 2010 at 10:32 am
Steve Thompson-454462 (4/2/2010)
As to TIS, if this requirement is being dropped doesn't the J - for Journeyman need to be dropped as well (in my mind this implies having completed some sort of apprenticeship, or level of experience). Just a semantic point, but the cert name might not make sense if you can achieve it fresh out of school.If the biggest problem with TIS is the overhead, but it still seems valuable for the validity of the certification, then, maybe it should stay but not be strongly enforced. I mean, wouldn't the cert panel be in the same position as any job interviewer: you ask for x years of experience, the applicant supplies a resume that documents that experience; during the interview they answer questions that demonstrate they have the experience they claim and if there's any doubt, references are checked.
If we divide knowledge into "book smarts" and "practical knowledge" (one you get from studying, one from solving real world problems), and the goal of this cert is to acknowledge applicants who can demonstrate both, than TIS may be important. But the dichotomy is not that sharp; i.e. the two types of knowledge reinforce each other.
Therefore, to "game" the TIS, an applicant might falsely claim experience that he or she does not have, but they can only get away with this if they can capably pass the other criteria (which I believe may include both a written and practical exam); therefore only candidates who are borderline might need their experience questioned and in these cases this could be verified by checking a random reference or two.
Candidates who are competent in the other requirements may be able to get away with fudging their experience, but in these cases you don't push it, since competence is competence. So basically, TIS becomes a "nominal" requirement (you're supposed to have it), but it's only going to be verified in the cases of borderline candidates (of course the applicants don't need to know this).
Wish I had seen this prior to my response. This makes a lot of sense in favor of TIS.
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 2, 2010 at 10:35 am
CirquedeSQLeil (4/2/2010)
You make a good argument against the Time in Service.
If we drop the time in service, would we implement more tests? I see more tests as a way of validating more increased knowledge about a topic.
I'm going to stop worrying about time in service because it seems like most people like it. It's the tests that are the more important topic anyway. The tests need to test both knowledge and ability. The point being that unlike MCTIP it's not simply saying "Yes, at this moment in time, this person knew this information." but instead "Yes, at this moment in time, this person demonstrated this knowledge and the ability to apply it." I think that will be more useful. It's the mechanics of administering such a test, in a way that people can't easily game it, that will be hard.
On the one hand we have automated mechanisms, say with virtual servers set up just such a way that people have to install, configure, document, troubleshoot, tune, the databases and code on them, and then an automated scoring mechanism... or an easy manual scoring mechanism. On the other you have people in front of a board, say MCM people or some other designated group, who will test the individual's knowledge and abilities on the spot and then, presumably, vote, or score the individual.
I think the latter offers the best chance at an ungamed system and, in a perfect world, the best chance to provide a really accurate measure that the individual testing really knows their stuff. But, the peer panel also offers some horrific chances for prejudice, picque or even indigestion to screw up the test so that Individual A is really certified, Individual B is not but passed, and Individual C should have passed but didn't. You'd almost have to videotape the reviews and have a second panel check the first, or put the individual testing behind a screen with voice masking technology or something so that any issues the testers had with Purple followers of Valen wouldn't influence the test.
In short, this ain't easy.
I'm leaning towards automating the technology because I think that offers the best chance to get this cert to a wider audience so that it's more useful for more employers & employees.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
April 2, 2010 at 10:37 am
Lynn Pettis (4/2/2010)
I think I can see people focusing this cert only to DBA's. I don't see that as the case. This should be certification that is also targeted to developers, and with a possible split, to BI developers as well. I split DBA/Developer and BI based more on the focus of the parts of the SQL Server product each use, not saying a developer won;t use SSIS or SSRS, but who is more likely to use various parts of SQL Server.I definately see a split in the MCJ; MCJ SQL Server and MCJ BI.
Absolutely. There needs to be no confusion on the matter and there must be a delineation between the competencies. Maybe with the delineation their could be an additional "emphasis" attached to the certification, i.e. MCJ BI emphasis SSAS.
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 2, 2010 at 10:41 am
Jack Corbett (4/2/2010)
To this point I've stayed out of the more detailed discussion, but I'll weigh in on the TIS.I think it needs to stay in. I agree that it would not be feasible, at least to start, to enforce before granting the certification, but I think that it should be included so that, if someone has been found to have lied about it, the certification can be revoked. In this business, being ethical is very important and any unethical behavior should invalidate a certification.
I am quite elated at these points Jack. You bring up two very important topics: ethics and revocation of certificate. In the past, the certificate holder could only be revoked his certification for violation of NDA (as far as I know). If we throw in this facet, I think it immediately increases the legitimacy of the cert.
We are charged with a very important job - Protect the company data. If one is unethical about gaining a certification, what does that say about what that person may do with the data. Maybe nothing, but it could be something for an employer to consider.
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 2, 2010 at 10:48 am
Kit G (4/2/2010)
Just a note on requiring Time in Service. What are you going to do if the company that the person was working for goes out of business and the upper executive structure goes to jail for the crimes that caused the business to fail?
You know, this scenario was exactly what I was thinking of when I read Steve's suggestion, and why I like it. The person says what experience he has, but the certifying board doesn't have to verify it. And, if it turns out that the person was lying, they can be de-certified. Yes, it would be easily gamed. Yes, it doesn't require verification. But the emphasis of this certification is the knowledge, not the TIS. The certification requires the experience, but the remainder of the process can weed out those that don't have the level of knowledge required.
Wayne
Microsoft Certified Master: SQL Server 2008
Author - SQL Server T-SQL Recipes
April 2, 2010 at 10:50 am
WayneS (4/2/2010)
dma-669038 (4/2/2010)
Am really sorry to hear Wayne thinks he is 'waffling' about it , I didnt think he was and i dont' think it should just be tossed out either.Hey, I didn't say I had changed my mind now. Just that Grant made several good arguments against it, good enough for me to question why I wanted the t-i-s.
Steve Jones - Editor (4/2/2010)
I think you ought to keep TIS, though perhaps make it lightweight to administer. Have someone list the time and percentage of that time, working with SQL Server. Developer might be 1 year * 50% of time writing code.People will game it, but it's just a rough measurement of experience and skill. After all, 10 years for a DBA can be 1x10 years or 10x1 year. Is that better or worse? Just have someone document it, and as far as employers go, if they choose to verify, or look at it against a resume, they can add their own weight to it.
Steve, I really like this idea - it keeps the TIS, but minimizes the administrative effort of it. Would you say the 1 yr * 50% = 1 yr experience, or 1/2yr experience? I would lean towards the 1yr myself.
This makes very good sense, and this method is part of resume' writing in several companies (the company resume' format, not the individual's). It is more about what the individual claims his competency and experience is versus what he/she is able to answer on tests/labs/interviews. That kind of evaluation would indeed be very valuable to an employer.
April 2, 2010 at 10:50 am
Steve Jones - Editor (4/2/2010)
I think you ought to keep TIS, though perhaps make it lightweight to administer. Have someone list the time and percentage of that time, working with SQL Server. Developer might be 1 year * 50% of time writing code.People will game it, but it's just a rough measurement of experience and skill. After all, 10 years for a DBA can be 1x10 years or 10x1 year. Is that better or worse? Just have someone document it, and as far as employers go, if they choose to verify, or look at it against a resume, they can add their own weight to it.
I think this is an important facet of the TIS. Documentation of the time one has spent working with the product. It should be lightweight, but is necessary for making the certification more appreciable.
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 2, 2010 at 10:51 am
Lynn Pettis (4/2/2010)
If you spend 50% of your time coding, what were you doing with the other 50%? There is more than coding, so I have to say 1 year is 1 year.
Yes - 1yr is 1yr.
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 2, 2010 at 10:53 am
Grant Fritchey (4/2/2010)
CirquedeSQLeil (4/2/2010)
You make a good argument against the Time in Service.
If we drop the time in service, would we implement more tests? I see more tests as a way of validating more increased knowledge about a topic.
I'm going to stop worrying about time in service because it seems like most people like it. It's the tests that are the more important topic anyway. The tests need to test both knowledge and ability. The point being that unlike MCTIP it's not simply saying "Yes, at this moment in time, this person knew this information." but instead "Yes, at this moment in time, this person demonstrated this knowledge and the ability to apply it." I think that will be more useful. It's the mechanics of administering such a test, in a way that people can't easily game it, that will be hard.
On the one hand we have automated mechanisms, say with virtual servers set up just such a way that people have to install, configure, document, troubleshoot, tune, the databases and code on them, and then an automated scoring mechanism... or an easy manual scoring mechanism. On the other you have people in front of a board, say MCM people or some other designated group, who will test the individual's knowledge and abilities on the spot and then, presumably, vote, or score the individual.
I think the latter offers the best chance at an ungamed system and, in a perfect world, the best chance to provide a really accurate measure that the individual testing really knows their stuff. But, the peer panel also offers some horrific chances for prejudice, picque or even indigestion to screw up the test so that Individual A is really certified, Individual B is not but passed, and Individual C should have passed but didn't. You'd almost have to videotape the reviews and have a second panel check the first, or put the individual testing behind a screen with voice masking technology or something so that any issues the testers had with Purple followers of Valen wouldn't influence the test.
In short, this ain't easy.
I'm leaning towards automating the technology because I think that offers the best chance to get this cert to a wider audience so that it's more useful for more employers & employees.
This is where I think Microsoft really comes into play. Currently, certification testing is done through testing centers, but with this certification, more robust systems may need to be in place, including things like virtual labs. Microsoft could host such an endeavour and hopefully provide the necessary support to develop the automated testing needed to make this work.
April 2, 2010 at 10:56 am
I do think we need to have automated test. I'd be in favor of a VM that is set to some base and we ask for you to accomplish xxx tasks. We then have an automated way to check that you did it. Maybe not how you did it, or how many mistakes, or even how long, but that you accomplished it in the time allowed.
April 2, 2010 at 10:57 am
CirquedeSQLeil (4/2/2010)
Lynn Pettis (4/2/2010)
If you spend 50% of your time coding, what were you doing with the other 50%? There is more than coding, so I have to say 1 year is 1 year.Yes - 1yr is 1yr.
I'd lean away from years. I'd actually go to months, especially as I expect many projects, or even engagements to last < 1 year.
I'd go towards gross months with SQL Server. Trying to limit it to subsystems would fail, IMHO. So 30% time as a developer over 2 years would be 6 months experience with SQL Server.
April 2, 2010 at 10:59 am
CirquedeSQLeil (4/2/2010)
Jack Corbett (4/2/2010)
To this point I've stayed out of the more detailed discussion, but I'll weigh in on the TIS.I think it needs to stay in. I agree that it would not be feasible, at least to start, to enforce before granting the certification, but I think that it should be included so that, if someone has been found to have lied about it, the certification can be revoked. In this business, being ethical is very important and any unethical behavior should invalidate a certification.
I am quite elated at these points Jack. You bring up two very important topics: ethics and revocation of certificate. In the past, the certificate holder could only be revoked his certification for violation of NDA (as far as I know). If we throw in this facet, I think it immediately increases the legitimacy of the cert.
We are charged with a very important job - Protect the company data. If one is unethical about gaining a certification, what does that say about what that person may do with the data. Maybe nothing, but it could be something for an employer to consider.
Ethics for the Data Professional is whole topic in and of itself, but I definitely think any certification offered by a professional organization should include agreement to a code of ethics and that said certification should be revoked if a person is proven to have violated said code. I believe that this lends greater weight to the certification. It won't stop unethical people from being certified but should help limit them.
A side note is that I have notified a member of the PASS board about this thread, they are excited about it, and may be weighing in on it.
Jack Corbett
Consultant - Straight Path Solutions
Check out these links on how to get faster and more accurate answers:
Forum Etiquette: How to post data/code on a forum to get the best help
Need an Answer? Actually, No ... You Need a Question
April 2, 2010 at 11:01 am
Grant Fritchey (4/2/2010)
It's the tests that are the more important topic anyway. The tests need to test both knowledge and ability. The point being that unlike MCTIP it's not simply saying "Yes, at this moment in time, this person knew this information." but instead "Yes, at this moment in time, this person demonstrated this knowledge and the ability to apply it." I think that will be more useful. It's the mechanics of administering such a test, in a way that people can't easily game it, that will be hard.On the one hand we have automated mechanisms, say with virtual servers set up just such a way that people have to install, configure, document, troubleshoot, tune, the databases and code on them, and then an automated scoring mechanism... or an easy manual scoring mechanism. On the other you have people in front of a board, say MCM people or some other designated group, who will test the individual's knowledge and abilities on the spot and then, presumably, vote, or score the individual.
I think the latter offers the best chance at an ungamed system and, in a perfect world, the best chance to provide a really accurate measure that the individual testing really knows their stuff. But, the peer panel also offers some horrific chances for prejudice, picque or even indigestion to screw up the test so that Individual A is really certified, Individual B is not but passed, and Individual C should have passed but didn't. You'd almost have to videotape the reviews and have a second panel check the first, or put the individual testing behind a screen with voice masking technology or something so that any issues the testers had with Purple followers of Valen wouldn't influence the test.
In short, this ain't easy.
I'm leaning towards automating the technology because I think that offers the best chance to get this cert to a wider audience so that it's more useful for more employers & employees.
Due to the review panel, I would lean toward the automated mechanism.
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 2, 2010 at 11:04 am
Steve Jones - Editor (4/2/2010)
CirquedeSQLeil (4/2/2010)
Lynn Pettis (4/2/2010)
If you spend 50% of your time coding, what were you doing with the other 50%? There is more than coding, so I have to say 1 year is 1 year.Yes - 1yr is 1yr.
I'd lean away from years. I'd actually go to months, especially as I expect many projects, or even engagements to last < 1 year.
I'd go towards gross months with SQL Server. Trying to limit it to subsystems would fail, IMHO. So 30% time as a developer over 2 years would be 6 months experience with SQL Server.
Okay, 30% developer/70% DBA. Would that take the 2 yrs back to 2 yrs?
Could make that 30% Developer/20% Data modeler/50% DBA, how would that work out?
This is where I'd say if you worked with SQL Server for 2 years or 12 years, it doesn't really matter other than you have been working with it. The real test comes with the testing, do you demonstrate the necessary knowledge and application of that knowledge to accomplish specific tasks.
April 2, 2010 at 11:07 am
As I thought about the labs, even if Microsoft actually hosts the MCJ testing systems, these would probably be best accessed from certified testing centers to help reduce the possibilities of others taking the tests for friends.
Of course, if this were found to happen, and both were certified, both would lose the cert.
Viewing 15 posts - 571 through 585 (of 685 total)
You must be logged in to reply to this topic. Login to reply