Granny's Programming Pearls
"inside of every large program is a small program struggling to get out"
JavaRanch.com/granny.jsp
The moose likes Certification Results and the fly likes SCJP--Passed with 78% Big Moose Saloon
  Search | Java FAQ | Recent Topics | Flagged Topics | Hot Topics | Zero Replies
Register / Login
JavaRanch » Java Forums » Certification » Certification Results
Bookmark "SCJP--Passed with 78%" Watch "SCJP--Passed with 78%" New topic
Author

SCJP--Passed with 78%

sever oon
Ranch Hand

Joined: Feb 08, 2004
Posts: 268
Hey all,
I guess this is the place to report this...I've been writing in Java since day one, so I pretty much expected to breeze through this test and I did. My test preparation and technique might be of interest to some, so:
I studied about 3 to 4 hours over the period of about a week before sitting for the exam (I scheduled it before I started). I read all the questions through without the intention of answering any. The only questions I did answer on the first pass were the ones I was extremely sure of. (Anything with an exhibit button, for example, I skipped on this pass.) I kept careful count of every question I answered with a high degree of confidence. Once I had the minimum required to pass, I answered the next 10 or 12 easiest questions just for good measure and guessed C on all the rest. At the end, I counted 46 answers I was positively sure of, and scored 48 right (I guess two guesses hit?). Total time in and out: 35 minutes (for the exam portion).
My initial thought was to spend as little time as possible by making sure I got the 31 and then guessing randomly, depending on the 20% or so guess hits to compensate for any wrong answers in that 31...but I chickened out of that plan and played it safe. (It's $150, after all. )
My impression of the test is that it should be relatively easy to pass for anyone who's programmed in Java for more than a year. I feel this is Sun's intention; this is based on the fact that they only require a 52% (31 right out of 61 questions), and random guessing nets you 20% on average (well, ok, a little less because some of the questions require more than one choice). My feeling is that the test would actually be more challenging if they made the questions a shade easier and raised the passing percentage to 70%--that way a test of "equal difficulty" would skew more towards the people that know Java and less towards the lucky guessers. Or, they could penalize 1/5th of a point for wrong answers? Just a thought.
One problem I did have with the test, as with most tests in Computer Science, is that many questions tested knowledge built into the compiler that is of little value to a programmer. Question: given this code, will an exception be raised or will compilation fail? Answer: who cares? That's what compilers are for. It's more important to be able to recognize that there *is* a problem than whether that problem will cause a compilation failure vs raise an exception. There are questions that fall into this category that are valid b/c knowing the answer makes clea some deeper truth about things...but I looked into the exam questions in this category with this in mind and was still not able to see this point on most of them.
Also, knowledge of JDK method signatures was a bit over-represented I feel. I don't know about you guys, but I've been writing Java in industry for several years now and I use the Math class so infrequently that I always check the javadocs when I use it. So I'm not clear on the point of memorizing whether Math.ceil(6.7) will return 6 versus 6.0 (not an actual exam question--don't freak out).
Another distinction that irked me for this certification (again, not on the test I took, but I ran across it in a cert prep book):

B "is-a" I? true/false
S "is-a" B? true/false
S "is-a" I? true/false
Most people I know would say that all are true. According to my review book, the SCJP exam only considers the second one true. "Is-a" apparently doesn't cover implementing interfaces in their view. This is silly to me--it's splitting semantic hairs that have little or no impact on how one designs a system...the interface establishes a means of referring to implementing classes polymorphically. Therefore, those subclasses can be treated as though they "are" of that type. No?
So in the grand scheme of things, these are mostly nitpicks. Overall, I thought it was a fair test and pretty much in line with what I expected, and I'm looking on to the next ones. The only question is...which one should I tackle next? Dev, business, or web?
sev
Satish Avadhanam
Ranch Hand

Joined: Aug 12, 2003
Posts: 697
Hi Sev, Congrats!!
Originally posted by sever oon:
Hey all,
So in the grand scheme of things, these are mostly nitpicks. Overall, I thought it was a fair test and pretty much in line with what I expected, and I'm looking on to the next ones. The only question is...which one should I tackle next? Dev, business, or web?
sev

Well, I think which one to take next depends on the field you are working. If you are more into presentation layer and working with JSP's or Servlets go for SCWCD(By the way, a newer version of exam is coming soon in a month or couple). If you are into business layer and dealing with EJB's daily, go for SCBCD. And you are developing desktop Java applications, SCJD is the best.
Hope this helps. Good Luck
George Thomson
Ranch Hand

Joined: Jul 09, 2003
Posts: 107
congrats,


George Thomson<br />SCJP, SCWCD, SCBCD, IBM-141(XML), SCEA<br />......................................................................
 
I agree. Here's the link: http://aspose.com/file-tools
 
subject: SCJP--Passed with 78%
 
Similar Threads
'Who wants to be a millionaire?' style practice test
Passed with 84 percent.
Passed SCJP 1.4 with 91%
another set of codes
Passing Strategy from Middling Student