lunarfollies

I think people have to remember that >50 for some people may be 0 and for others it may be 45 that is a huge difference. The AM part of the Exam was by far the most difficult part to study for and was also my lowest scoring by a long shot.

my $.02 , on what worked for me.

I put in allot of practice on the AM, I had many of the "canned answers" memorized and could write out everything very fast. I memorized all of the key language in the guidline answers, "full credit" answers that Schweser gave. I worked hard on writing all of my Port Mgt answers in a consistent easy to follow and understand spreadsheet format similar to how CFAI does it. Remember the graders look at the guideline and then look at your answer the easier to read and closer it looks to guideline the more partial credit you get.

Many of the old exam answers I wrote over and over again knowing if I ever got a question on a similar organization or person I would be able to bang out the answer. I figured this may give me time for a curveball question or to review and I did finish almost 1 hour early. This also got my arm in shape my hand was not very tired during the exam.

I brought in a blue and black pen (graders use red, don't ever use red it pisses them off) I did everything in black but I circled all answers in Blue. When I did a formula I used arrows in blue to label every input, I made sure every step was clear. I also went back and made sure every answer was clearly marked.

Looking at my scores every part that I scored >50% for a fact I wrote down the formula and or RRLLTU, etc and calculated something (obviously wrong) but I must have decent credit not zero.
I had 3> 50% and 4<70% but I'm pretty sure I got 100% on 2 or 3 of those questions so in some cases the essay helps.

chunty

janakisri Wrote:
-------------------------------------------------------
> Doesn't say much about the program and the quality
> of the test if the absolute majority scores low in
> most of the essay q's . Just reeks of ambiguous
> wording and subjectivity. My PM looked fabulous ,
> but the AM was a disaster. How is that possible?


I think there's a fairly straightforward explanation for the low-AM-score / higher-PM-score phenomenon:

In the PM portion, L3 is all multiple choice, so the first 33% of a candidate's score is essentially a freebie: even someone that has never heard of the CFA curriculum should score 33% just by guessing. From there, it doesn't take too much learning (or even luck / random variation) to reach >50%.

In the AM portion on the other hand, most of the questions are written answers, so there are far fewer freebie grades: an unprepared candidate is starting much closer to 0%, and is much more likely to fall short of the 50% mark.

If you look at the score matrices of passing candidates posted on AF, you'll see that for the most part, the distributions of morning scores fairly closely resembles those of afternoon scores.

soverby

Sorry but your argument doesn't work because your key assumption for scoring 33% (i.e. random guessing) can not be used for a candidate who, instead of randomly guessing, is trying to select the correct answer.

That is, if you randomly guess, your expected return is 33%. However, if you instead attempt to answer multiple choice questions based on a combination of the provided information and your knowledge, this 33% is no longer relevant and your 'starting point' is back at zero.

BC_MBA_student

SFA Wrote:
-------------------------------------------------------
> That is, if you randomly guess, your expected
> return is 33%. However, if you instead attempt to
> answer multiple choice questions based on a
> combination of the provided information and your
> knowledge, this 33% is no longer relevant and your
> 'starting point' is back at zero.


I disagree.

Say that I'm an economist, and each day I need to use available information and my knowledge to predict whether the USD will A) Appreciate, or B) Depreciate versus the Euro.

I'd expect an unknowledgable person to be able to predict this correctly on 50% of days.

By your argument, the professional economist should be measured against a benchmark of 0% correct -- so if he predicts correctly on 25% of days he is a good economist.

I don't think that's right. You need to compare the economist's performance against what one could achieve (50%) by knowing nothing and guessing.

Similarly, on a multiple choice test with three choices, a 33% score should be considered the baseline; that's what performance must be compared against.

John10

It is unfortunately not possible to distinguish between the questions you "know you know" and the ones you "know you don't know". I think this is the crux of the issue.



Edited 1 time(s). Last edit at Saturday, August 20, 2011 at 01:21AM by SFA.

Otabek

My point is that in all of these situations, there is a freebie baseline, that doesn't suddenly disappear if you're "trying" rather than guessing.

Another example:

The return performance of a large-cap PASSIVE portfolio manager is expected to be on-par with the return of the S&P 500. The passive manager is not trying to beat this expected performance.

A large-cap ACTIVE manager on the other hand is trying to add value -- in other words, he's "trying" to select superior securities. Should we benchmark him against a 0% return because he's "trying"? Nope. The performance of the S&P is still the baseline; and he only is considered a knowledgable/skilled portfolio manager to the extent that he beats the S&P.

To summarize: same baseline applies, whether you're accepting the expected baseline performance, or trying to beat it.

IAmNeil

look, this would work if you could approach each question asking "do I know this 100% for sure?" If yes, then select correct answer. If no, randomly guess. Then you would be correct. But the reality is that we are never 100% sure that we know the correct answer. There is always at least some doubt. Now also consider the case where questions are designed to create additional ambiguity and doubt.

Your perspective is a 'law of large numbers' perspective. I.e. take enough random trials and you will approach 1/n. What I am talking about on the otherhand and what I think is reality for CFA candidates is that for each individual question if we don't have the material down COLD we will more than likely fall for a 'distractor' option and score zero for that question. Also, the correct answer will unlikely present itself as an attractive option to those that are less than fully prepared so for the 'prepared' candidate the chances of unwittingly CHOOSING (not guessing) the correct answer is very small. And that is to say NOT RANDOM.



Edited 1 time(s). Last edit at Saturday, August 20, 2011 at 02:35AM by SFA.

lucasg85

The problem is that you end up in a circular argument which I think we all have nightmares about... "I think the answer is A, but I am not sure. Since I am not sure, A is likely to be a distractor, therefore the answer must be B or C. What do I do? Random guess: ER = 33%. But if my paranoid reasoning is correct picking B or C yields a ER = 50%. But what if my hunch of A is correct?"

This I hope illustrates the difference between randomly guessing and having some idea what you are doing. While random guessing should get you close to 33%, having some idea of what you are doing could well get you 0%.

ruchita

speaking of economics, have you guys noticed that stuff (e.g. burritos) keep getting smaller and crappier? Could this be disguised inflation (i.e. instead of price going up, products just keep getting smaller and crappier)?