Beta Results *should* be available by late April

[Bert Bates]

Hey guys,
I just talked to Evelyn. She *thinks* that the SCMAD beta results should be available in about two weeks - roughly the end of April.
Please, please, please - do NOT bet your bottom dollar on this estimate - and don't bug Evelyn - this is just an educated guess - as it gets closer we'll have better information!
Bert

[Sivasundaram Umapathy]

Thanks for the info bert
Keeping my fingers crossed for the results.

[Ken Ho]

Thanks a lot for the update.
Keeping nervous of the result.... :roll:

[Jim Crawford]

Who should I bribe to pass?
I pay top dollar.
Just kidding.

[Yi Meng]

hi,
I saw SCMAD on sun website that it will be available 23/04/04 with the following info: (http://training.sun.com/US/catalog/courses/CX-310-110.html )

Delivered at: Authorized Worldwide Prometric Testing Centers
Prerequisites: Sun Certified Programmer for the Java 2 Platform (any edition)
Exam type: Multiple choice, drag and drop
Number of questions: 68
Pass score: 55% (38 of 68 questions)
Time limit: 150 minutes

Does this mean that all beta results are finished marking and out? Why do we still have to wait???

[Forrest Xu]

Passed score is 55%!
So for beta test, the passed score should be kept in the same, right?

[Nart]

Hello,
Yes the beta passing score will then be 55% as well. You are marked as if it was the actual exam.
Since its released on the 23rd, then it really seems that all marking is over, then maybe the results should be out before Bert suggested
But thats only a MAYBE!
Anyways the passing score of 55% lifted my hopes up a bit, but i still doubt i'll pass this one :roll:
Best of luck to all
Nart

[Eduardo Rodrigues]

I prefer don't pass to accept such ridiculous pass score!
Disapointed again...
Why the Sun's certifications are so easy to pass???
Bye

[Mark Spritzler]

Originally posted by Eduardo Rodrigues:
I prefer don't pass to accept such ridiculous pass score!
Disapointed again...
Why the Sun's certifications are so easy to pass???
Bye

I don't know about you, but I know a bit about J2ME, and thought that I would score my highest on this exam, since this time I had a bit of experience backing me up.
But, sorry, I did not do well on this exam, and I really thought I knew the material really well. It just didn't ask questions in the areas I thought it was going to ask. So the test was really difficult, I don't think I passed. But not because I can't develop J2ME applications. I can and understand them pretty well. So I think based on the toughness of the test, only the best will pass. So 55% is not low in that respect.
Mark

[Eduardo Rodrigues]

Hello Mark...
I really think that the level of questions are very good... I don't think that I will pass in this beta too, but...
The point here is: 55% is very low!
Maybe in one or two months there will be a lot of SCMAD... Worst , they won't be recognized as they deserves...
Only my two cents...
Bye

[Ken Ho]

It's really surprised that the passing score is 55%.
At first, I think it would be around 60-62%.
Also, Sun made relevant changes on the certification learning path diagram. A pretty cool view.
[ April 17, 2004: Message edited by: Ken Ho ]

[Nicholas Cheung]

In fact, passing rate is not that important, the important point is that you can get high score in the exam, and thus, that's the difference between 2 SCMAD holders!
In addition, you can show your score reports to the interviewers when you are in the interviews.
However, I regard certificates as a bonus, not a must. Thus, I will only take exams only if the cost is low or even free.
It is seldom that the employers will give us a job solely becos you have some certificates. Especially, for example, you will not employ by SUN solely becos you got all SUN's certificates with 90+ marks. But if you work in SUN already, having such certs. may give you some advantages over others.
Nick

[Nicholas Cheung]

Also, Sun made relevant changes on the certification learning path diagram. A pretty cool view.

SUN will add one more after May.
Nick

[Yi Meng]

Originally posted by Nicholas Cheung:

SUN will add one more after May.
Nick

I just wander how does SUN put the web services in......J2EE? not really......

[Bert Bates]

Just curious for those of you who think that 55% is too low a passing rate, what percentage of candidates do you think should pass the test?

[Ko Ko Naing]

Well, the most exciting time is closer and closer... At the end of April!

[Mike Wang]

Originally posted by Bert Bates:
Just curious for those of you who think that 55% is too low a passing rate, what percentage of candidates do you think should pass the test?

60-70% passing rate

[Bert Bates]

Mike, et.al,
Let me rephrase my question...
If 100 candidates take the test, how many candidates do you think should pass? Forget about the grade, just think about the percentages of passers and failers...
Bert

[Jason Hocker]

Bert, with the passing score set at where it is, what percentage of test takers passed? I understand if this is a guarded secret, but it may help us to understand the process better.

[Bert Bates]

Jay -
See? That's just what I'm getting at! So I'm putting it out there to you guys first - If 100 beta candidates take the beta, how many should pass?

[Cleverson Schmidt]

See? That's just what I'm getting at! So I'm putting it out there to you guys first - If 100 beta candidates take the beta, how many should pass?

My guess: 20

[Sivasundaram Umapathy]

20? too low
Bert, pls tell us the approximate number of passed candidates for 100 candidates

[Eduardo Rodrigues]

I think 70%...

[Mark Spritzler]

Originally posted by Eduardo Rodrigues:
I think 70%...

You think 70% of the test takers should pass? I think around 50%.
Mark

[Amy Caruso]

Great then, 60% may be optimal?

[Bert Bates]

So, while Sun's formula is more complicated than this, a big part of the calculation is that X% of the beta candidates will pass. I don't know what this percentage is, but let's say, just for an example, that Sun says that roughly 70% of the candidates should pass. If that was the only criteria (which it isn't), then, theoretically, if the exam was really tough, the passing percentage could be 20%!
The point is, that in general, an exam's passing percentage is really meaningless. The real point is what percentage of candidates pass - if 90% pass, it's an easy test regardless of the passing percentage, and if 20% pass it's a hard test regardless of the passing percentage.

[Eduardo Rodrigues]

Hi Bert
But as everyone knows, some candidates (maybe the majority) take the beta only for try, best effort or whatever, and some candidates (majority again) don't even know about what the subjects of the exam really means...
For example, when I pass in the SCBCD exam (beta), I never read nothing about EJB's up to that time... And I can manage the test to pass... Sure, I learned A LOT in the process, maybe this was the best part...
What I really want to say is: maybe Sun may promote some kind of pre-beta candidates selection, to get only people who knows what about the test is...
Maybe in this way the pass percentage will be greater and the title gained will be more valuable on the market...
Just my 2 cents
:-)
Bye

[Nicholas Cheung]

However, for example, if you and I applied for the beta, how SUN can determine which one is eligible to take the test, while the other cannot?
Let say I have no experiences in J2ME, and you have 2 years experiences, by what mean SUN knows? I can cliam I got 3 years experiences as well. If there is a pre-test (or entry test), then SUN may need to manage one more test before the real test.
If SUN only sends us some questionaires for simple test, I can always refer to the books or articles, etc. Thus, it is very difficult to carry out a pre-test, if not impossible.
Nick

[Eduardo Rodrigues]

Hi Nicholas!
I show the question but I don't know the answer...
But , IMHO, with the low pass percentage the Sun's certifications are losing the value on the market ...
I agree, the most important benefit is the knowledge acquired by the process of studying , but it will be wonderful if the certification itself means "Hey, man, this is THE java guy!"... Don't you agree?
Best regards,

[Sathya Srinivasan]

Hi,
I would agree with Bert in that the exam pass percentage doesn't really matter. If I understand correctly, the pass percentage is based on a statistical analysis of the beta results and hence has been based on the relative difficulty of the test according to the test takers.
However, I do have some thoughts on that.
This exam is slightly different in that the area is relatively new and not too many people are actually working on it (at least from what I can observe from the posts by test takers).
This would mean that the beta exam would be perceived as a tough exam. However, as time goes, I would expect J2ME (hopefully) to gain more mass and that people would starting working in J2ME, thereby developing practical knowledge. Once that happens, the exam would not look as tough as the beta testers felt and hence might pass easily, thereby breaking the statistical analysis. Lack of proper exam resources for beta test takers also adds to this problem.
One more observation I can make from this thread is that there is a perceived value to a pass percentage. Most of us, regardless of which country we are in (I studied in India and in USA), have developed a notion that 70% - 80% and above (A or B grade) is a good score. To suddenly see 55% as a pass score (equivalent to a D) creates a psychological feeling that the exam is not strict enough.
I can see two ways out of this.
1. Sun should revise the pass percentage on a periodic basis so that future test takers find passing the exam relatively difficult. I believe the exam versioning mechanism, although a money-making scheme, also helps alleviate this problem indirectly.
2. Sun should provide a grade along with the score and make it mandatory to display a score when people say the passed a certification, much like standard exam. This can potentially be controversial since people who pass the exam with, say, 60% would not be happy to tout that they got a C on the exam. But it will definitely be fair to those who passed the exam with 95%
Just my 2c.
Sathya.

[Theodore Casser]

Originally posted by Sathya Srinivasan:
I can see two ways out of this.
1. Sun should revise the pass percentage on a periodic basis so that future test takers find passing the exam relatively difficult. I believe the exam versioning mechanism, although a money-making scheme, also helps alleviate this problem indirectly.

I think this is actually a good idea, if anyone's asking opinions. It keeps the exam at a relative difficulty given the market - in other words, as more people have greater skill with the technology, it remains a yardstick of general qualification.

2. Sun should provide a grade along with the score and make it mandatory to display a score when people say the passed a certification, much like standard exam. This can potentially be controversial since people who pass the exam with, say, 60% would not be happy to tout that they got a C on the exam. But it will definitely be fair to those who passed the exam with 95%

I can't agree with this one, sorry.
There's just too much wrapped up with requiring a grade or the score to be coderanch. There're plenty of folks who know the material who just don't test well, and the score doesn't reflect how well they know the material. There's those of us who figure out how many right answers we need to pass and then call it quits after that point. And there's the fact that if you continue to use the skills that were tested, the score is no longer an accurate reflection of ability down the road (which would mandate period retests to show this, etc...)
Nah. I think the current system (just the cert) is there for a reason.

[Sathya Srinivasan]

This can potentially be controversial

Hence my disclaimer!
The first option would definitely be less controversial, but has its own logistic problems. Sun has to advertise the revised pass percentage constantly, which can be a headache. But, for the $150 that they charge, I think it should be ok.
Also, it will give an excuse for people that they did not know what the pass percentage was!!
I agree with a couple of points raised that a grade does not reflect the improvments one makes after taking the exam. But, then again, it is the same for college degrees also.

[Jim Crawford]

Basically I don't feel I deserve to pass because I know how much I knew of the subject matter (please don't fail me just for saying that ;-) ).
I got 90% on SCJP and I still felt very 'unready' to take the exam - it says something about my opinion on the exam standard.
I thing the following nails the problem and provides good solutions.

Originally posted by Sathya Srinivasan:
... the beta exam would be perceived as a tough exam. However, as time goes, I would expect J2ME (hopefully) to gain more mass and that people would starting working in J2ME, thereby developing practical knowledge. Once that happens, the exam would not look as tough as the beta testers felt and hence might pass easily, thereby breaking the statistical analysis.

One more observation I can make from this thread is that there is a perceived value to a pass percentage. Most of us, regardless of which country we are in (I studied in India and in USA), have developed a notion that 70% - 80% and above (A or B grade) is a good score. To suddenly see 55% as a pass score (equivalent to a D) creates a psychological feeling that the exam is not strict enough.

This is also very true and is basically what bothers many people. Maybe the actual question content should be mathematically manipulated to 'insert' more easy questions in order to achieve a higher passing score.

1. Sun should revise the pass percentage on a periodic basis so that future test takers find passing the exam relatively difficult. I believe the exam versioning mechanism, although a money-making scheme, also helps alleviate this problem indirectly.
2. Sun should provide a grade along with the score and make it mandatory to display a score when people say the passed a certification, much like standard exam. This can potentially be controversial since people who pass the exam with, say, 60% would not be happy to tout that they got a C on the exam. But it will definitely be fair to those who passed the exam with 95%

I think these are both good options and I think you should show your score. If the score bothers you, retake the exam.
Bert - comment? To unrealistic to be implemented right? Doesn't make enough business political sense right? Oh well...
Cheers.

[Bert Bates]

A bit of wild speculation on my part...
I'll ask Evelyn if she might consider watching the exam for a couple of months, and perhaps consider raising the passing percentage a bit once more materials are available. My guess is that this would probably run into some legal snags. For sure, Sun factors in the survey results when setting the passing rate.

[Nicholas Cheung]

I think the low passing rate is expected for new exams, such as SCMAD and SCDJWS.
It is becos there may not be so many experienced MAD and JWSD to take the exam. Thus, if SUN wanna 60 people out of 100 pass the exam, SUN orders the marks in desending order, and the 60th people's mark is the passing rate.
Since we are new to these technolgoies and with a limited time to study, our marks maybe low, and thus, say, the 60th people's mark is 55%, and SUN uses this as the passing rate.
In fact, the passing rate reflects the beta testers' ability on this area, and thus, it is REAL. However, I saw some of us think that the passing rate is too low, and ask for a revise.
Personally, the beta test is used to determine the real passing rate, and now the rate is REALLY that low. If SUN tries to adjust the rate, then SUN in fact does not carry out the beta test. SUN can simply say, no matter how people perform, the passing rate is 65% or 70%.
As a result, if most people are in the range of 60%-65%, most of them fail. But just SUN does not want to do this.
Nick

[Theodore Casser]

Originally posted by Bert Bates:
A bit of wild speculation on my part...
I'll ask Evelyn if she might consider watching the exam for a couple of months, and perhaps consider raising the passing percentage a bit once more materials are available. My guess is that this would probably run into some legal snags. For sure, Sun factors in the survey results when setting the passing rate.

I would think the only real legal problem would be if the marker moved. Specifically...
If it moved up, you'd have to deal with possibly revoking certifications. I know that I, for one, would be fairly upset to get a letter from Sun telling me that they've raised the grade and now my prior one doesn't measure up to the level required. (I'm not expecting that to happen, in that I'm not expecting Sun to request materials back.)
If it's moved down... I think you enter into an interesting situation where folks who previously failed are going to demand to be measured at the new standard.
Damned if you do, and damned if you don't. I don't envy having to make decisions on these matters, much less on what passing ought to be in the first place.

[Jim Crawford]

I guess with a thread like
"Anyone failed?"
either noone posts a fail, or (too) many passed.

...
2 cents.