FTR, this is the Modified Angoff Standard Setting Model: https://www.questionmark.com/us/seminars/Documents/webinar_angoff_handout_may_2012.pdf
Just an FYI-- the phrase “95% sure” doesn’t really have a meaning in statistics (it is a confidence interval). The word “confidence” has a specific meaning that “sure” does not (even though you’ll see people use it like it’s the same) 
…I’d also be interested in your methodology here.
This being a forum, I’m using considerably more relaxed language 
Plus, it’s not a true 95% confidence interval because I wasn’t entirely sure how to combine the 3 aspects of it.
I calculated a mean of 86% on my “sure” questions, 61% for “unsure”, and 34% on guesses, with sample standard dev of 6.6%, 9.3%, and 21% for each. I had 11 tests so that was my sample size. I initially used the t distribution but then decided that was overkill. So I set my confidence interval to -3 standard deviations and +1.8 standard deviations to be on the conservative side. So obviously this is not a true confidence interval, but I know that I can be at least 95% sure that a more thorough confidence interval wouldn’t be worse than the one I computed.
Here’s my problem: if I were to assume these were proper confidence intervals, how do I combine them to get one overall confidence interval for my whole mark? I’m at least 95% confident my “sures” were better than 66%, but I can’t just take the bottom limit for the sures, unsures and guesses, because really that would mean that I’m in the 5% worst case scenario for ALL 3, which is extremely unlikely. That would give me a lower bound of 50%, and it’s damn near impossible my mark was that low. What I did was assume a scenario where each of the sures, unsures, and guesses were at their lower bound, and the other two were at their average (so 3 scenarios), then I took a weighted average of these. That got me the 64% mark. Then I did the same using the +1.8 z score and got the upper mark of 82%.
I’m not aware of how to take 3 confidence intervals and have some sort of weighted average of them into a single confidence interval. I’m pretty sure my method is fairly reasonable… But if you have any advice on errors I’m making, please give me some advice.
Edit: Oh and I was “sure” on 73 questions, “unsure” on 40 questions, and “guessed” on 7 questions. This ratio is a fair amount better than any of the mocks I took (other than BSAS, which was way too easy) and my lowest mock mark was 63%.
Oh my God. You were being serious. Sparetime, if I failed I was going to hit Tickersu up for some stats tips and tricks, but you man, you’re going to be my coach who creates oodles of spreadsheets to measure my progress. I could get a friend who’s a published author write a book on The Making of CFA Candidate, or something compelling like that that might sell on Amazon for $9.99
^ go for it lammy, I would buy the book … Haha.
Oh I have not recovered from the studying and the exam… I still feel so tired and worn out… I sleep like a log.
I was fairly sure on 71 questions (59%, questions I didn’t have any doubt about). Giving a 25% (lower % to account for the few questions I may have missed in the first group) to the remaining 49, I should arrive at 83/120, therefore 69%.
I estimate a score of 82/120…hope the MPS is below 70%.
Last year I was band 9 and I’m fairly sure I scored around 56-58%, so my guess is the mps was near 65%.
@BobbyBraveheart Can you possibly share your scorecard? Your score shud have been higher than 58%, bcoz consensus is tht 2014 MPS was around 80/120. From Band 9 to Pass, there cant be difference of more than 5-6 ques at most.
I quite agree with you Love_Lord. I had a Band 8 last year, and my scorecard was: (Ethics and FRA > 70), (Alt Inv, Corp Finance, Derivatives, Equity, and Quants 51-70), and ( Eco, Fixed Income, and Portfolio Management <=50). Like I said, this was a Band 8 score, so I don’t agree with Braveheart’s assertion of 58% for a Band 9.
I scored Band 9 last Year. Scorecard was
FRA, Alt, Corp >70
Equity, Ethics, Port, Economics 51-70
Derivatives, FI, Quant <50
This year I feel I did better from last year but not sure whether it will be lead to a Pass. Would sadly be another close call this time 
That’s Level I question 
Q# Topic Max Pts <=50% 51%-70% >70% - Alternative Investments 18 * - - - Corporate Finance 36 - * - - Derivatives 36 - * - - Economics 18 * - - - Equity Investments 72 - - * - Ethical & Professional Standards 36 - - * - Financial Reporting & Analysis 72 - * - - Fixed Income Investments 18 - - * - Portfolio Management 36 - * - - Quantitative Methods 18 - * -
58% was my guess based on how I scored on mocks last year, I’ll never really know the true score of course
Last year’s MPS was very high… Assuming I was the lowest in all my ranges… I scored around a 64… and its not likely the lowest possible in every range i was in…
I think it was like 66-67 last year…
I was a band 10er
Last year I was above 70 in EQ , Accoutning, Corp, and Quant
50-70 in Econ, Fixed Income, Derivatives, PM
Fail Ethics, Alternatives
The matrix in post #54 could have correct answers in the range of 57-78. With only one matrix no conclusion can be drawn, It needs many border line matrix to find the narrow overlapping range of passing and not passing. Also, the guy has high ethics score and that could have helped him.
.
.
^ ProjectPat901. You are seriously funny. LMAO.