I don’t get your calculations. You’re using a binomial probability distribution I’m guessing? Why do you have 84.5 instead of 70? And at the end, this is just a test of significance. The exam will still be hard, but the mean will DEFINITELY be pushed up. Your calculations above are suffice to show the reasoning: Mean for 3 options: 1/3 * 120 = 40 Mean for 4 options: 1/4 * 120 = 30 That’s a 10 questions difference. The standard deviations are not very different (5.16 vs. 4.74). So the entire normal distribution curve will shift slightly to the right. I might have missed something, please point it out if I have.
I’m using DeMoivre-Laplace http://en.wikipedia.org/wiki/Theorem_of_de_Moivre-Laplace with 120 trials. So 70%*120 = 84 - but you need to adjust for the Normal distribution being continuous - as opposed to the binomial which is discrete. That gives you 84.5 The mean will definitely pushed up, but the probability of someone being able to pass through random chance will still be vanishingly small.
Its a level playing field, everyone is taking the same exam and the cream always rises…if you are as prepared for the next exam as you were for the last exam you can expect the same results…luck favors the prepared, not the guessers!!!
It will definitely be harder as the choices will resemble eachother so much. Another good way to rack in some more cash on the CFAI’s part.
BondClipper Wrote: ------------------------------------------------------- > I mean, theoretically a monkey walks in there now > and picks “A” for every answer, and scores 25%, > more than a third of the way to a passing grade. The monkey will score 1/3 now instead of 1/4… an increase of ~8.3% in absolute terms or ~33.3% in relative terms.
what if they made it like the SAT…with a penalty for wrong answers (that would encourage people to skip questions instead of guessing).
that would be nasty indeed - but nothing on the cfai website/faq about this change mentioned that so assume that guessing is still not penalized. Can assume they will compensate for the small guessing gain with harder questions and higher passing threshhold
I thought this was a complete joke when I first saw this thread… WOW
this is going to be very tough…you will now have 3 answers that all seem viable, rather than CFAI setting a question with 4 answers where 2 of them you can discount almost immediately if you don’t know the stuff inside out, then you will suffer here with this change
??? why when it will be 3 options all 3 will be hard, yet when its 4 only 2 are hard?? its a level 2 CFA = always hard whether you prepare or not!
I predict they will make the numeric answers more tighter around the true answer, and the qualitative answers more subtler around the true answer. And they will make the 2 wrong answers the most common results of faulty processes that candidates would be expected to make. In fact, expect that - I doubt they would make the alternative answers that far apart from each other anymore. They would use this change to reward more precision in thinking and calculation. Even more reason to strive to pass this as soon as possible!!! The rising candidate numbers globally are making them progressively raise the performance threshold while saving them time/effort/money.
Is it for L1 or L2?
all levels from what one could reasonably interpret from their faq release
Here my 2 cents… I really dont like this idea. first of all, even if the chance of passed with pure guessing is almost 0%, the chance for bad educated guesser will significantly rise. In my opinion educated guessing posses also important informations about your knowledge and skills, since a broad knowledge is the fundament for this. I dont think, that CFAI will raise the 70% line. so if the scores will rise higher than CFAI was expecting, the pass rate will also rise significantly, what will lower the level of quality of the CFA designation and is especially unfair to passes before 2009! When i read through the reasoning fo the CFAI, the passage – “In evaluating CFA exams we found that the fourth choice on the majority of selected response questions (multiple choice and item set) is frequently ineffective – attracting relatively few candidates” – sounds to me, that CFAI is hoping to attract more candidate by the psychologie, that “it is kind of easy to pass with only 3 choices”. If this would be the case (–> overall quality of candidates decrease) and CFAI would maintain the passing rate this would mean, the quality of people who passes will decrease… What do you think, do I interpret the passage right, I am not completely sure last, 3 choices could not apply really good the answers like “correct incorrect”, etc afte all this negative things, I still think there is something positive in it. since 3 choices would increase to overall %-level of Questions answered right and I dont think CFAI will increase the 70% pass level(as they dont say anything about this), it could be supposed that the recent MPS is well below 70%. If the MPS would be e.g. 68%, the 3 choices would significantly increase the pass rate, which I am sure the CFAI doesn’t want to!
I read the above as CFA saying that very few candidates pick one of the 4 answerw on any given question. If I understand correctly they’re saying: A) Correct B) The trick option C) Shows you don’t understnad the topic sufficiently D) Also shows you don’t understnad the topic sufficinelty but less likely to be picked than C on anverage (for whatever reason). By taking away D they still leave the trick and a random wrong answer so they’re is no point to D. I don’t agree but think that is their logic. As others have said the main change will be how they cope with 2 column answers as yes / no for 2 questions won’t work.
With the 3 answer format, they might choose to use K-type questions; that are those kinds of answer options that goes like this: A) II and III only B) II and IV only C) I and IV only and make I and III really tricky and plausible and similar variants - that’s one way for them to get around the 2 column/yes-no questions that are moot under the 3 option format.
At least thay won’t be able to make the double questions with: A yes yes B yes no C no no D no yes
Yes, I know that was their past policy - with the change, perhaps they might revisit that. We will see how they handle this when they issue the new texts and practice problems this fall. Of course, this only makes more work for the review providers who now have to re-do lots of practice problems and rethink old ones to adhere to the new format. CFAI and the review providers are in symbiosis - and the changes drive more business to them as candidates see value in doing prep ahead of the test.
I dunno about the symbiotic relationship between prep provders and CFAI - wzupdok (cute, btw) and I agree on that, but I don’t think CFAI does. Anyway, my thoughts: a) Probabilistically I don’t think this makes much difference. Randomly guessing among three choices or four is pretty much pointless because (1/3)^n goes to 0 remarkably fast just like (1/4)^n. If you are aiming to finally beat the 30% barrier, this change will help you alot. b) I think that the CFAI analysis of looking at the distribution of answers for test questions and finding that only a small percentage picked, say, response D so we should just eliminate choice D is pretty dumb. That says that the test designer can look at the choices and find the one least plausible to candidates before they take the exam. That’s hard to do with your own students using the text you picked to say nothing of a multi-national group using all kinds of diverse learning methods. c) Four is really convenient for lots of standard format questions - the ‘true; true’, etc questions. Those questions are just going to look odd. d) Test prep provders have huge banks of questions that are really helpful. Forcing them to redesign those questions just doesn’t help anybody. No matter what CFAI thinks, Schweser, et. al. is good for them. Why CFAI wants to degrade the available study materials for their own exam is beyond me. e) The argument that somehow three choices instead of four make the exam harder because you can eliminate two out of four but none out of three is astonishingly dumb. f) Having three choices instead of four screams out that the exam has become easier even if that isn’t true. You just can’t win an argument with someone who says “Three choices means that you get more points for free” by claiming “Well, psychometrically,…”. It sounds like blowing smoke. That means that CFAI is going to give people the perception that the exam has become easier and that means they risk losing the value of he credential. Risking that seems incredibly dumb. g) And on this “Psychometrically,…” topic… So CFAI has decided that three choices is somehow better than four because some psychometrician told them so? How many psychometricians work for CFAI? It’s funny how ETS employs tons of psychometricians but all their multiple choice exams have four choices instead of three. CFAI thinks they know more about test design than ETS? Anyway, here’s a really radical idea - why not allow the test deisgners to put as many choices for the answer as seems suitable for the question? What is gained by having the same number of choices for each question? Maybe they would get the best test by allowing test designers the freedom to make some questions “true/false” and some questions A - E.
I honestly couldn’t agree more. feel like sending this to Mr. Diermeier directly.