CT6 April 2011

[scaron]

So what are the views on this sitting? I thought it was a really interesting exam. Really enjoyed the twist on the GLM question as well as the lengthy decision theory question or how about the run off triangle question? What fun we actuarial students get to have

 

[Viki2010]

so did everybody finished everything in time?

 

[scaron]

On a serious note, is there no one out there that feels cheated out of those long hours spent revising for this exam? I personally think that this was the hardest ct6 exam I’ve laid eyes on, not just the technical aspect but in terms of time limitations. I can’t be 100% certain of this and the thought has crossed my mind that it could just be an overreaction due to a poor performance on my part. I had hoped that my somewhat tongue in cheek response would elicit a deluge of responses telling me that I was a (fill in the blank). After this exam I really am revaluating whether the sacrifices are worth it. This is the first time that it felt like the whole process was an utter waste of time (except of course when I had to learn the nonsense that is “exposed to risk” in ct4). Is there really no one out there that shares my grief?

 

[mdsath]

It's difficult to detect tone on the internet. I thought you were simply an overenthusiastic student pleased to be given a "challenge" so I decided to retreat back to the comfort of my own room hoping to never visit this forum again until results day (didn't last long).

My friend and I both shared your opinion. It was harder than the average CT6 paper, it was difficult to finish in 3 hours (I suspect only a very small minority would have) and there were lots of strange questions. I wouldn't go as far as saying I feel cheated but I can see where you're coming from. When I first scanned the run off triangle question, I assumed the little table near the bottom was inflation rates for accident years and we would simply have to repeat the method adjusting for inflation - how wrong I was. Just when I thought I was beginning to successfully fudge my way through the acceptance-rejection method, they give us a double exponential function. The question about calculating the expected profits [6 marks] and then the probabilities [8 marks] for each arrangement was "brutal" as my friend put it. Strangely, the time series question was one of the tamest there!

I'm still slightly optimistic though. I feel I've achieved a borderline score and so the pass mark will be crucial.

 

[maz1987]

Surely no one can deny it was harder than previous CT6 exams - even those that finished it (not that I was anywhere near finishing it). Almost every question was a curveball. I have tons of queries regarding each question, but I guess I'll save those.

 

[maz1987]

I'm still slightly optimistic though. I feel I've achieved a borderline score and so the pass mark will be crucial.

That's where I sit. I feel I demonstrated enough to show that I know the subject to a decent level. I did not wing my way through each question, and hope that that's enough to pass.

 

[maz1987]

After this exam I really am revaluating whether the sacrifices are worth it. This is the first time that it felt like the whole process was an utter waste of time (except of course when I had to learn the nonsense that is “exposed to risk” in ct4). Is there really no one out there that shares my grief?

It's bound to happen more than once in anything you do. Don't give up!

 

[beyers]

During the exam I skipped all questions where I struggled to immediately see the solution. That saved some time, but at some stage I skipped three questions in a row and thought I am in great trouble ...

In the end I revisited those parts. I was also in time trouble, but when I looked at the paper afterwards I felt that it went much better than it felt during the exam.

I agree that the exam was quite a bit more difficult than the average test over many years - however it was possible to get easy full marks for the first 30 or so marks.

In hindsight I think that the most difficult aspects were:
1) Time problem
2) The acceptance-rejection method question
3) The last 6 marks regarding lognormal distribution and development factors
4) Compound distributions part where we need a normal approximation

 

[John Lee]

Well just sat the paper under exam conditions and I have to say it was a difficult paper (took me 2hrs 10mins instead of the usual 1hr 40mins). This was caused by a higher number of higher order (ie unusual) questions which included a number of derivations rather than application.

But hey! The time series was easy...even if it did have a mistake in it...

 

[Viki2010]

Well just sat the paper under exam conditions and I have to say it was a difficult paper (took me 2hrs 10mins instead of the usual 1hr 40mins). This was caused by a higher number of higher order (ie unusual) questions which included a number of derivations rather than application.

But hey! The time series was easy...even if it did have a mistake in it...

And what mistake did you detect?

 

[John Lee]

And what mistake did you detect?

It was a repeat of the ARIMA mistake in S05 Q8 - you can't say it's ARIMA until you check it's stationary - whereas they did this in part (ii).

 

[DevonMatthews]

Well just sat the paper under exam conditions and I have to say it was a difficult paper (took me 2hrs 10mins instead of the usual 1hr 40mins). This was caused by a higher number of higher order (ie unusual) questions which included a number of derivations rather than application.

But hey! The time series was easy...even if it did have a mistake in it...

You can usually do a full 3 hr exam in under half the time? That is phenomenal....

 

[maz1987]

To be fair if you are very knowledgeable at a CT module, which I hope the tutors are, then a standard exam could be completed within 1.5 hours. It's reassuring that a tutor found this exam to be significantly harder than the others!

 

[DevonMatthews]

To be fair if you are very knowledgeable at a CT module, which I hope the tutors are, then a standard exam could be completed within 1.5 hours. It's reassuring that a tutor found this exam to be significantly harder than the others!

Knowledge has very little to do with fully completing some exams in the allowed time. I'd like to see someone do some of the harder CT5 exams in 1.5 hours giving full and complete answers.. They take 4-5 hours for a capable student to complete.

 

[scaron]

I don't entirely agree with you DevonMatthews, I think that it depends on which CT subject we are considering. In the case of CT5 you are entirely correct.


But in the specific case of CT6, I think Knowledge has a lot to do with it! If you work on CT6 for years going over all possible questions, setting questions and explaining it to others then you should be expected to complete the exams much quicker than say someone who has spent a relatively small amount of time on the subject. Lots of practice allows one to have to think less, know the appropriate shortcuts and to generally answer questions much quicker. As my tutor at university would say, "..an exam is no time to be thinking about questions. Every question on the exam paper should seem as though you've already answered it before".


I think a suitable analogy is that of a pupil, at primary school, who is impressed with their maths teacher who knows their times tables inside out and seems a whizz. It's only later on, once you yourself become proficient, that you wonder why the primary school teacher was happy to show off their times table skills to 7 year olds who were only just starting to master it themselves.

 

[maz1987]

Knowledge has very little to do with fully completing some exams in the allowed time. I'd like to see someone do some of the harder CT5 exams in 1.5 hours giving full and complete answers.. They take 4-5 hours for a capable student to complete.

True, it has little to do with some exams, so I take that back. But the exams such as CT1, CT6, CT7 etc can be completed reasonably quickly if you know the material. Not just familiar with it, but know it inside out .

 

[DevonMatthews]

It depends on the actual exam and the questions on them. It's possible to construct questions (especially in CT8) that even experienced tutors would struggle with, and it's also possible to construct questions that are excessively computationally intensive. You could easily make a CT4 question that is within the syllabus and asks to calculate the 11 step transition matrix of a system with 6 states... Knowing the course inside out wouldn't make the calculations any faster. There are numerous questions like this that have appeared in past exams before.

 

[scaron]

It's reassuring that a tutor found this exam to be significantly harder than the others!

I totally agree with you on this. Hopefully this will lead to a reasonable pass mark being set.

Oh and thanks for your positive comments maz1987.

 

[maz1987]

If the tutors are allowed to speculate, may I ask what may be the reason for the Institute setting an exam like this that is significantly different to previous exams - as this one obviously was?

 

[scaron]

Though I’m not a tutor, I speculate that the reason for such exams is to ensure that at least 45%-60% of students who sit any given exam are likely to fail. This is because;

a) If too many students pass then this would take away from the “prestige” of being an actuary
b) If too many students pass it would dent the image of actuarial work being hard
c) Students that take tests that are tricky and impossible to complete in the allotted time will continue to perpetuate the myth that actuarial work requires intelligence and is itself difficult
d) Contentious point: This is a business and students who fail have to repeat the exam (and possibly have to purchase updates or take private tuition) ensuring that profits are made all round

When we end up debating whether understanding and knowledge of a subject are enough to complete an exam or if we can agree that some exams would require 4-5 hours to be completed by capable students then something surely has to be wrong with the whole exam setup. I find that there is something farcical about the whole exam system and it seems to me that no one, for whatever reason, is able or willing to debate openly and expose this nonsense process for what it is.

(It feels so good to finally say that out loud and in public)

 

[mattt78]

4-5 hours

When we end up debating whether understanding and knowledge of a subject are enough to complete an exam or if we can agree that some exams would require 4-5 hours to be completed by capable students then something surely has to be wrong with the whole exam setup. I find that there is something farcical about the whole exam system and it seems to me that no one, for whatever reason, is able or willing to debate openly and expose this nonsense process for what it is.

(It feels so good to finally say that out loud and in public)

In theory the standard required to pass an exam should be much the same from sitting to sitting, so in that case it doesn't really matter if its a relatively easy or hard exam (as the pass mark should vary to compensate), the more important question is if its a fair exam - i.e. a true test of knowledge and understanding of the material. Arguably a difficult exam will more easily distinguish the better candidate, and therefore be fairer. However, fluctuating pass rates suggest the required standard to pass is not consistent, and this is where the examiners lose credibility in my opinion.

Of course, if an exam really takes a good well prepared candidate 4-5 hours to do, it may end up testing exam technique as much as anything else, which may not be a very fair test. (But to be honest, I find it difficult to believe the 4-5 hours figure)