I read a commentary that stated no one has ever failed Level II who scored above 62%, not sure how credible it was but there was certainly a basis behind it, can’t seem to dig up the article though
There is no way it is 82%, not sure where that info is coming from…I’m talking about the lower bound of the range. You saying 82% is like me saying, well, no one whoever scored a 95% has failed, obviously 82% = autopass, that goes unsaid.
Of course you can get below 50% on a section. Just make sure your overall score >70%. You’re taking Level 2, surely you can run the numbers to see how many questions you can get wrong. Remember the sections can contain a varying number of questions. I suggest you run a sensitivity analysis.
The one section you should never get <50% on is ethics. Just don’t do that.
Hey, I actually posted on this a few weeks back when I returned from my CFA review class. It was the Kaplan three day crash course and the instructor, Dr. Spieler, brought this exact point up. He has been teaching these review classes for 10 years as well as having his grad assitants research results. There are some years where the L2 passing score has been a 61%. He said if you can get to averaging 4/6 per item set you should be fine. People can debate this all day but it would simply be their fear and speculation versus ten years of experience and research.
And thanks kd26gioi for sharing that bit of info…actually it helps to ease the pressure a bit by knowing a realistic estimate of the passing score rather than people making generalized statements like score>90 and you pass!! or going through threads like getting 70+ consistently…
Should aim to get high scores but a realistic estimate helps!!
I don’t know how credible this is. But if this is true, and 61% has been the passing rate, it means people are probably only getting 50% before the random guessing adjustment boosts their score. That says to me that the exams are so difficult now, and the scores are low enough, that CFAI literally holds total control in determining the number of people that pass or fail each exam.
And taking it one step further, I bet pass rates may be lowered in the future…
I’ve been thinking about the idea of minimum passing score with randomness and quessing. Let’s say CFAI wants people to get 70% right on the exam for a pass (this could be any number). They also know that people do guess and they’ll have to compensate for it. We know that, if you know you will at least get 66 questions right, and guessed on the other questions, you’ll get around 70% - if everything is random. So, there is a confidence interval for 70%, it is set at 55% (66/120) + the lower bound of the 95% confidence interval of 18 correct with 1/3 randomness.
No need to calculate this, I just ran a monte carlo simulation with 100,000 runs. It’s around 11 questions for the lower bound of 18 questions correct. (the standard deviation of 18 correct questions is 3.46)
So, this means the minimum passing score would probably be around 64% (66+11=77, 77/120=64.17%) if they want people to pass with 70%
And yes, I know CFAI has their modified angoff, and people taking the exam to test questions, and the final say is up to them, but all of that aside, this IS the number.
No prob, dude! Now, how the frak am I gunna score 64%? I don’t know,
Although, I don’t think CFAI is this sophisticated to apply random guessing and a confidence interval when setting the pass %. It’ll probably be some intern clicking the AVERAGE function on excel to come up with the number.