Nov 2001 Exam verdicts:

[Retired Bookworm]

As the exams complete, what is everyone's verdict on the difficulty, fairness etc. of the exam?

Course 4: SoA stuck to the habit of throwing every oddball credibility question they could muster, but surprisingly they overlooked numerous regression questions, which almost looked like Actex problems. Rounding is still a terrible nightmare in the exam, when 90% accuracy narrows down the choices to 40% of the choices (ranges between answers on a later question were separated by $.02 each, I guess an auditor wrote this one). Overall, the exam was easier to trek through than May.

[Anonymous]

So, what do you think the pass mark will be? What was it in May, around 22?

[RTVA]

I think pass mark was 22 on C4 in May, and 23 on C3.

[Retired Bookworm]

I didn't see significant changes in difficulty to the May exam, which would suggest we'll see another 22 to pass.

[New at pd]

Hi everyone,
After taking a power nap after the exam, I realized that this exam was much like the May '01 exam. Lots of slam dunks, several curve balls, and some "What the hell are you saying?" questions. Hope that 22 works as a passmark.

[Anonymous]

After recovering from a severe depression after the exam, I realized people who pass Course 4 in 2000 are so lucky. Both 2001 exams are evil.

[M.]

I agree with Donkey Kong. The 2001 exams are inspired by none other than Satan. I thought they were trying to make Exam 4 easier. Instead, they seem to be compensating for the tiny reduction in material by making the questions a good bit nastier on average than the 2000 exams. I say, bring back the broader syllabus and give us easier questions.

[Anonymous]

Course 4 example #5 I cpouldn't figure out . they asked for a 90% conf. interval for Beta Iused the t dist with 6 d.f. ess/6/sum X^2.

[Lloyd Dobler]

That should have worked. Did you remember to take the square root?

[NYC Triathlete]

Course 4: I thought it was fair. Out of all 40 questions there were only 2-3 that I had absolutely no clue about. Unfortunately my brain froze on a couple of the Buhlmann credibility questions and I could not get an answer that was one of the choices to save my life. Also--I did not appreciate all the "read these 5 statements and pick out the false one" questions since I detest questions like that. My "favorite" was the one that had something about a BLUE estimator. I was surprised that there were no kernel questions, only one F-test question, and in general the regression questions were sort of out there. But anyway, even though they put a different spin on it, all in all it wasn't so bad. And I sure do hope you guys are right about 22-23 being the passing grade :smile:

[Anonymous]

My initial reaction is that this exam was slightly easier than 5/01 and much more difficult than either of the 2000 exams.

I'll wait until Rodger Lee's site is up an running before I make a final decision.

There were several questions that I said to myself, "They can't be asking a question this easy?!" I'm guessing that they weren't!

[erickson]

I'm going to go slightly against popular opinion here. I thought today's exam was more difficult than both Nov 2000 and May 2001. I took Nov 00 and May 01 each as practice, scoring 29 and 32, respectively.

But, I'm thinking that I'll be fortunate to get 24 on today's. I answered 28 confidently, (but that usually translates to about 80% accuracy) and had to take blind guesses on the rest (some due to time constraints).

There were too many "which one is false/true" questions -- I prefer calculating something and seeing that my solution matches one of the choices. Also, the Buhlmann Cred questions seemed less straight forward than those of prior exams.

[mememe]

I thought this exam was harder than the sample exams. I wish they had asked more straightforward survival questions. I got a couple of them but didn't feel that comfortable with the rest. I'm surprised that they keep giving plug and chug time series questions. I memorized the formulas but is that really testing my knowledge?

Do you think they adjust the difficulty/pass mark of an exam depending on the % of people passing in the previous session's exam. ie Whoa, we passed 40% this session. We'd better make the next exam even harder to push the pass % into the 30's.

<font size=-1>[ This Message was edited by: mememe on 2001-11-05 19:27 ]</font>

[Oscar]

I took all three past exams in the last 2 weeks, and I thought this one was the most difficult. Like others, I was tripped up by the Buhlman questions - a topic I thought I knew well. I defineately did not have my "A" game, and poorly worded B.C. ?'s aggravated this. I felt like I was kicked in the package. Normally my score reflects my preparation, but not this time. I'm praying for a 21 pass score. Every time I think of a problem, I realize i missed it. i left having answered 24, but now I'm certain a few were wrong. I say 18+guesses.
Who cares about me..sorry.

[Anonymous]

Have to agree w/ Donkey Kong. What the f*** are they doing? Ridiculously difficult imho.

[VP]

Add one more name to the 'evil test writer' chorus. I think there has been a noticeable upgrade in difficulty with each exam since May 2000. The most significant jump is 501 to 1101 but 1100 to 501 was pretty brutal too. I didn't like the 6 true-false types - was expecting one or two. I missed the kernel - that would have been in my gimme category. The only BC one that busted me, I think, was the "you observe at least one claim" Yikes. Spent about 25 minutes trying to get that empirically. Never used the normal table either!

[Bullseye]

I think I passed Exam 4. Obviously, I will have to wait 2 months to find out. I thought I was going to get a 5. I knew some stuff very well. And, I knew some stuff NOT at all !!! However, I am an expert Buhlmann credibility guy and they asked a bunch of questions on Buhlmann credibility. I answered 21 with confidence. I answered another 5 with reasonable confidence. The rest EEEEEEEEEEEEEE ! Good Luck Everyone!

[Course 5 Zoombie]

this exam really hard is all i can said.

[Anonymous]

I also thought the exam was ridiculously difficult. This is my THIRD time sitting for Course 4 after scoring a 5 and a 3 in November and May. I was totally overprepared for this exam, but, like many of you, got tripped up on a couple of questions and ended up spending too much time on those. That left me with two minutes of time and 6 questions that hadn't been done. A quick glance at those last few questions told me that they put some of the more straightforward questions at the end of the test this time - that totally stinks. I have yet to understand why it is that the CAS and SOA are determined to make these exams as difficult as they possibly can - what is the point?
If they're going to give problems where you have to read half a page several times over before understanding the question, then they need to give us unlimited time! I think that would be a true test of ability - not this crap where you have to work as quickly as you possibly can to get the problems finished.
I'm obviously a bit disgruntled over this exam!

<font size=-1>[ This Message was edited by: Pita on 2001-11-06 10:44 ]</font>

[Oscar]

Did this exam separate those that new the applicable material form those that did not? That is the question. I always contend if you know it well enough, it will show. This time I'm not so sure.
Generally the most difficult prblems on all 4 exams were around numbers 21-35. I always jump to the end after number 20, or whenever it gets harder.

[Anonymous]

Course 4 was rough. What did you guys get for the right truncated..like .40 or E?
Anyone get .86 for a Buhlmann one?
How about the log rank one..I got 3.4 but I regret it now.

[Anonymous]

I do recall getting a .86 for a Buhlmann problem. Hope we nailed that. GL w/ the results, everybody.