They are open book with no proctoring (video).
They do however run anti-cheating software which does have some issues.
They are open book with no proctoring (video).
They do however run anti-cheating software which does have some issues.
There is no reason exams shouldn’t be open book. Give students the source material. We should already be testing application over rote memorization anyway. Make it more like the real world where we have access to the information at our fingertips and test the understanding of the fundamentals.
I especially didn’t understand why the EA exams weren’t open book. We all have copies of the code and the regs, I referenced them weekly if not daily in my job, but yeah in order to be a great pension actuary I apparently need to know some obscure rules about a 4044 allocation.
I have lobbied for open book CAS exams for years. But that’s a big change, and not something they can just do administratively.
CIA has a P&c track now with open book exams. Hopefully more Canadian companies start supporting this track instead, since CAS has no jurisdiction at all in Canada. I would switch to CIA track, but I am to deep into CAS exams to make it worthwhile.
Any bets on the pass rates this sitting? 50% is probably the absolute minimum across the board, most likely in the 60-70+% range.
I’d say around 65%
For which exam?
Exam 9 had 60% last sitting for the overall rate. I thought that was crazy high. 63% if you go by the effective rate.
Was just speaking generally, not a specific exam in particular. Honestly that wasn’t too high for exam 9 standards, but I expect we’ll see an even higher pass rate for Exam 9 this time, probably record high.
Wouldn’t surprise me if all the exams end up around ~70+% range this time as a form of damage control.
You’d think it might not be too hard to make really minor changes. Change some numbers on the computation questions or change a detail on essay questions that wouldn’t materially change the difficulty but would be something different.
Candidates who had seen all or most of the exam would know what to study, but at least it’s not like they could walk in knowing “the first question is 20.7 years and the fourth question is $792.50 a month” or whatever.
Maybe not every question, but tweaking some would be better than tweaking none.
It’s still not ideal, of course, but it’s better than ignoring the issue.
I suppose that’s better than what the CAS used to prefer to do in the event of a bad exam: set a high pass mark / low pass ratio, and blame the students for being unprepared.
i’d second this. if the exams have a MC answer system it’s more important. if they are free response then the actual work is on the set up and process, but still some tweaks should be pursued.
as to overall fairness, I’m, not concerned about the perceived fairness. there is no truly equitable outcome once the vendor crashed within the testing window
They are open response. I doubt it’s actually helpful to remember “the answer i got was 4.62”. I doubt tweaking the numbers would make much difference to the difficulty for anyone who has a chance to attempt the problems before.
It would be a lot more valuable for me to know they tested topic x in paper y and added a specific twist than to know the answer was 123.45
I remember back when I was writing exams, doing the exam post-mortem, where I considered some questions (usually with others at my company, since this was before the days of the AO or GoA) and realized, with clarity made possible by not being in the crucible of the exam room…or because my coworkers were more clever than I… the trick to answering the question that I missed during the exam.
Having advance knowledge of which tricks are needed this sitting is a definite advantage. Once you know the trick…arithmetic is easy (especially when the answer sheet is a tab in an Excel-like spreadsheet).
Yeah, changing the numbers so the answer is 54.321 rather than 123.45 only does so much, for sure.
Knowing the tricks, knowing what to study is far more valuable.
But changing that is hard. Coming up with a whole new exam of equal difficulty level hard.
And there is value in knowing that the answer is 123.45. If you know the question is coming and you get 246.80 but know it should be 123.45 then you know you screwed something up and need to fix it. So there’s some value in going after the low-hanging fruit and changing the numbers.
Tweaking the questions to make slight changes is a LOT easier and still has SOME value.
You’re certainly not going to achieve perfection; we’re looking for a reasonable place to land on the fairness vs feasibility spectrum.
Also, I’ll just reiterate my view that the exams are a form of hazing and we’ve become so incredibly jaded that we view something that is only even attempted by a tiny group of incredibly intelligent people who spent hundreds of hours in preparation as being “weak” or “easy” if 80% of them passed it.
Which is a preposterous standard if you take about 20 steps back and look at it objectively.
96% of first time medical board takers pass and that seems to be ok with most people.
So with that said, if you know what to study, and study it and pass the exam, that sounds to me suspiciously like… learning.
When I was sitting for exams, due to the various issues with CAS exams, I was convinced that the exams were more about limiting the supply of actuaries than they were about confirming the qualifications of actuaries.
I thought then, back when online learning was just starting to be explored, that the CAS would be better served by having a sequence of online classes to teach the required material, complete with homework, interim assessments, and final exams instead of the one-shot, all-or-nothing self-study exams.
With time and experience (and the CAS discovering that some material is better covered in “modules” than exams), I’ve relaxed that view somewhat. We want actuaries to be folks who can learn new complex concepts with self-study, so a demonstration of being capable of such is appropriate in the credentialling process.
But I do think that moving away from self-study for everything and implementing common CAS-sponsored classes for a few more topics would still be an improvement from the hazing process we still have. And it would open a door to involvement with universities, making use of their educators’ skills at crafting theoretically consistent, quality exams… (although a couple of my university professors did have some horrendously lousy weed-out exams).
Actuarial orgs definitely function as a guild (they do restrict supply via exams to some extent in order to keep wages higher for their members)
This can also go wrong eventually, like it has for accountants in the US.
Accountant shortage prompts US plan for quicker path to qualification - Accountant shortage prompts US plan for quicker path to qualification via @FT