[jft]

Apparently, not too many people really know what a CPA exam is. I have given up explaining to them what it is exactly so I just tell them its harder than the bar exam. I think I read one person on a blog who said this. if you have taken both, can you please confirm which one is harder?

FAR- DONE

BEC-DONE

AUD-5/14

[How many letters do you need]

CA bar has roughly a 30% pass rate. Some people take it once and are so beat up by the process never try again…regardless of the fact that they just spent 100k+ on law school. The structure of the exam itself is more rigerous….The volume of information is much more to know at one time. I can take FAR and then forget most of it if I want to because my next exam won't have more than a 10-20% overlap at most. I think “harder” is a tough question to answer because some people are naturally geared toward being a lawyer…I'm sure the bar is easier and likewise the same for people who have natural talents in accounting. But from a strict structure/pass rate perspective I don't see how the CPA is nearly as difficult.

MBA,CMA,CPA, CFF?, ABV?

[Anonymous]

I have a friend that I used to work with who has taken and passed the bar exam and also taken and passed the CPA exam. I had asked him which he though was harder and he said in his opinion “the CPA exam was far more difficult”. Although I think there are so many factors involved that this would vary depending on who you asked.

@ How many letters The bar exam pass rate in CA is actually 41% right now.

https://focusbarreview.com/bar-exam-pass-rates-by-state.html

My friend took the bar in MA and pass rate here is 63% which is higher than pass rates for the CPA exam in 2014 which is on a cum between 48% and 56% depending on which exam. All pretty interesting.

[StevenFrost]

How many letters do you need and CPA_Driven

It is difficult to compare the pass rate of one exam to four because in order to become a cpa you must pass all for.

Your probabilities only account for the probability of passing one exam. What is the probability of passing all four on your first try?

Let's assume that the probability of passing regulation, financial, business, and aud individually is around 48 percent.

The probably of passing all four exams on your first try then would be, I believe, correct me if I'm wrong…the result of .48*.48*.48*.48 = .05.

So the probability of you passing all four exams on your first try, roughly estimated, is around 5%. It is probably much lower than that however, since I have accounted for the issue in a very general sense.

I believe this calculation is parallel to the calculation for getting heads (a passing mark) four times in a row, since each previous exam is not necessarily correlated to any additional bonus for passing the second..

https://answers.yahoo.com/question/index?qid=20090408114106AAg7nTA

Is a good answer too…getting heads four times in a row has a 6% chance of occurring, around there….This is assuming that a person who does the bare minimum of study but more than enough to pass has a 50 percent of achieving a 75…or something of the matter.

Which of course is why the actually probability of passing the exam is probably more around 1%…on your first try, since it's a matter of putting in the work….not a matter of random chance.

In terms of pass rates then…we are comparing a CPA exam pass rate of 1%-10% for all four exams on the first try to a 30-80% pass rate on exams on the first try.

[RIST]

OP: I work with two Lawyer CPAs.

They say the bar can be beaten with a 1 month prep class even with a weak foundation and the CPA isn't just harder, the two shouldn't be compared. The bar is just an expected part of becoming a lawyer while the CPA exam is a brutal test to get rid of the weak and highlight the strong.

[JamesBJames]

@StevenFrost, passing all the exams on the first try isn't going to be as low as 1%. You might want to brush up on your statistics. The CPA exam is hard, but it cannot be that hard.

If all the exams are independent from each other, then the probability of passing all four is roughly .5^4 or 6.25%, sure.

This assumption is weak in two areas, however. Firstly, a person is more likely to pass the exam the first time than he is on subsequent attempts (and NASBA has data on this). I forget the exact statistics — it's something like a 58% chance of passing any exam on your first try, but it dips to 43% or so for re-takes. Using .58 instead of .5 in the above equation gives a little over 11%.

Of course, the exams also aren't independent from each other. You can model this with Bayesian updating — if a person passes one exam, he is statistically more likely to pass the others. (e.g., I did great on FAR and AUD; BEC has a couple of topics that involve FAR and AUD -> I'll probably do better on BEC than I would have had I not taken FAR and AUD first)

I'm a very boring person, so I've looked at the variability in scores based on average test score. The trend is that the higher the average test score, the less total variance there is and the less overall co-variance exists between any two individual scores. This should raise the all-four-first-time pass rate above 11%, although I'm not sure by how much. Personally, I'm guessing it's somewhere between 15-20%.

Of course, that still makes it harder than the Bar exam if you use pass rates as the metric to compare the two. I would also assume people, on average, spend longer studying for the CPA exam.

EDIT: I just looked at that site linked above. Holy crap — an 85% pass rate for New Mexico? 80% for Florida? A lot of states between 70-79%? It'd be nice to take that test instead.

FAR: May 1st, 2014 - 91
AUD: May 29th, 2014 - 97!
BEC: July 16th, 2014 - 91
REG: August 29th, 2014 - 88

Licensed December 2015

Feel free to add me on LinkedIn by clicking my username!

[StevenFrost]

JamesBJames

I like your post. Please bare in mind that I already calculated that exact amount you stated if you read above my post…but I would giver a 1-10% range to be fair, not 20%-25%.

I think you need to go even deeper into the probability model you have listed.

First, there is the group of people who will be more likely to pass the second, third, and fourth if they pass the first. This is still in the minority. In an office I worked at I was able to obtain a sample of at least 25 accountants. Of those 25 accountants, only 2 passed all the other parts the first time after passing the first part. The other 25, of which passed through the office, I know their story, did not pass the exams with more ease.

So I think we should adjust your high percentage of 20%. I think your assumptions might be biased based on your own experiences. The experience of others has lead me to believe that passing your first exam does not make the other exams easier – although for me it did…for most it does not…based on my objective measurements or sample size, which is sufficient to project out with only a 5% confidence interval to the entire population of Chicago, since I am not only reporting on this group of 25 people but also the 100's of others who I know who still have not passed.

So to recap, James, your model fails because it assumes 100 percent of people who pass the first exam will have an easier time of passing the second, third, and fourth.

This however is simply not true. Passing the first exam does not make the second, third, or fourth easier except for a minority of people.

For most people, only passing the first AND second exam makes the remaining two easier. That's a conjunction James. That implies that both have to be true in order for the probability to increase, not one. In other words it is not sufficient a test taker pass his first exam to be more likely to pass the others, it is necessary for him to at least pass the first two both before he or she enjoys such bonuses.

But again, I already stated this in my post above, which you failed to read.

You see where your model fails?

Good try though. Stick to accounting, leave the actuary work to the ones with expertise.

See, you are correct in assuming that Bayesian updating will work here, which I also allude to in my post as a bonus percentage or probability, but you are incorrect in assuming that the Bayesian updating is linear. It is not linear, it is tiered and split between more groups of people.

[ScarletKnightCPA]

@How many letters do you need

Firstly,

If you want to compare by pass rates, Bar exam pass rate on average is far higher than 30% and it looks like California pass rate is higher itself.

Secondly,

The CPA exam is 4 sections taking individually. A more effective compare would be first time pass rates for all sections of the CPA exam, which is if I recall correctly, 20%.

Far: 76 (Wiley Test Bank)
Aud: 77 (Wiley Test Bank)
Reg: 61, 76 (Wiley book, Wiley Test Bank)
Bec: 86 (Wiley Test Bank)

MBA in progress

[ScarletKnightCPA]

@How many letters do you need

Firstly,

If you want to compare by pass rates, Bar exam pass rate on average is far higher than 30% and it looks like California pass rate is higher itself.

Secondly,

The CPA exam is 4 sections taking individually. A more effective compare would be first time pass rates for all sections of the CPA exam, which is if I recall correctly, 20%.

Far: 76 (Wiley Test Bank)
Aud: 77 (Wiley Test Bank)
Reg: 61, 76 (Wiley book, Wiley Test Bank)
Bec: 86 (Wiley Test Bank)

MBA in progress

[StevenFrost]

That;'s what we just said above…lol!

[JamesBJames]

I apologize if my post came across as aggressive, but tone down your attitude a little. This is an interesting topic to discuss. There aren't a lot of people who care that much about statistics.

I guess the main area we disagree on is the correlation. Your sample size and results suggest that there isn't much correlation between passing one exam and passing the other exams. To be fair, my sample size is probably biased as well — I used people I know in real life and people who have posted their scores in this forum, and I can only imagine some people don't self-report failing scores — but I'm reporting what I got based on my sample. Are you saying there is no correlation?

Regardless, I'm curious why you would put the long-run pass rate at anything below .5^4 or .48^4 or whatever number you use to represent the pass rates if you assume that they are independent from one another. (EDIT: Oh wait, you're giving a confidence interval. Never mind. Guess we're just arguing about correlation)

(Your note about tiered Bayesian updating is neat, by the way. I didn't consider that.)

FAR: May 1st, 2014 - 91
AUD: May 29th, 2014 - 97!
BEC: July 16th, 2014 - 91
REG: August 29th, 2014 - 88

Licensed December 2015

Feel free to add me on LinkedIn by clicking my username!

[StevenFrost]

JamesBJames…my apologies sir. I shall take it down a notch. I too felt a little aggressiveness but thank you for that acknowledgement.

Blessings to you brother!

The fact that you used the term Bayesian updating was very enlightening and helped me tremendously in understanding an idea I didn't have words for either for not remembering or never learning proper!

Thank you! This is a very interesting topic indeed.

[JamesBJames]

No problem, @StevenFrost. When you said you assumed the pass rate was between 1%-10%, I initially took it to mean that you thought it was a uniform distribution and not a binominal one. I think that made me more skeptical than I should have been.

I found this link to be interesting, by the way: https://nasba.org/files/2014/02/2013-Overall.pdf. If you're interested in…well, way too much data to handle, https://nasba.org/files/2012/09/2009-Main-Book-RS.pdf is pretty neat, too.

That said, I was a little off regarding the first-time pass rates. In 2013, the overall first-time pass rate per section was 54.6%. If you're being conservative and assuming no correlation at all between tests, then that gives just less than 9% as the population mean. If you apply your confidence interval, that gets somewhere between 4.5% and 13.5%.

We can disagree on the correlation; honestly, that's fine. I'll admit that getting a number that proves anything would require a ton of sampling data. I think I had 35-40 peoples' full scores when I did my analysis. I remember I got a small-to-medium positive correlation, but like I said before, there's some possible bias in non-reporting.

I kinda wish I still had access to the data I used, but I got rid of the file when I was cleaning my computer. I might have to get some new data and calculate P(passing all 4 exams | passing one exam first try), P(passing all 4 exams | passing two exams first try), etc.

Cheers.

FAR: May 1st, 2014 - 91
AUD: May 29th, 2014 - 97!
BEC: July 16th, 2014 - 91
REG: August 29th, 2014 - 88

Licensed December 2015

Feel free to add me on LinkedIn by clicking my username!

[StevenFrost]

Yeah James! I think that would be more accurate. The reason being is from studying I've kind of discovered that auditing and far are heavily interrelated…and so is BEC and AUDITING…but reg is kind of on its own…so it could also matter which one you take first.

For instance, if someone passes FAR or REG on their first try they would probably pass the other ones sooner than if they took BEC or AUDITING first. I know people who take auditing first almost always have to take one more again but could still generally do well.

That's from my observations here though.

[Vlakmir]

My roommate is studying for the BAR as I'm studying for the CPA this summer.

I started studying pretty much fill time, April 1st, I take my last test July 16th. 3.5 months of solid studying.

He started studying The last week of may. So The CPA exam takes at LEAST 1.5-2 months longer to cover the materials.

The worst part is, we have had discussions similar to this.

No lawyer that hasn't taken the CPA will ever admit to the CPA exam being harder. They paid $100,000 and were fed tons of bullcrap about being prestigious for the last 3 years, they -can't- acknowledge the difficulty because most people in law school are arrogant and think they are smart, despite being the joke of “smartest of the dumbest clumb (ie; English majors with A's that couldn't do sciences/engineering/etc…)

In the end, I think the CPA exam is harder, at least for my state's bar exam. It seems like the CPA candidates do a lot more work, and that in the end justifies my opinion. No one coming out of law school is any “smarter” than kids coming out of a masters in accountancy – they just studied different things, and clearly the CPA exam is more rigorous. Anyone who thinks a brand new law student is “smarter” than a brand new masters of accountancy student has just drank too much of the kool-aid. I mean, hell, it was $100,000 kool-aid

In the end, I think it's all just a pissing contest. Nobody should really care.

The world outside of law has the perception that law is prestigious due to the nature of work, TV, the media, and big $$ with corporate lawyers.

On the other hand, accounting has the stigma that it's bean counting and spreadsheets, despite the fact that it is most closely related to business law + Regulations + a quantitative output for all of those laws/regulations. However, lawyers are groomed from the start of law school to tell them they are important, meaningful, etc…there is a lot of “prestige” related to the profession due to the media/TV/people that don't know any better.

Guess which of these two are being sold to the young for high tuition costs to cover the costs for their universities undergrad programs….? Guess which has a horrible job market…? Yeah……

In the end, I feel sorry for lawyers. You come out of school with tons of debt, unfounded arrogance, and most likely not a job in a place you want to be / in an area you want to be.

My roommate studying the BAR

“omg I don't want to study about the statute of frauds……”

“You should use mnemonics? $500+ in writing, indebtedness of another etc..”

“no, no. You don't understand…like, all of my questions are really hard and complex, they can't just be memorized easily for short quick accounting problems”

me *sigh. Dumbass* “so isn't a mnemonic…exactly the reason….. yeah okay. Cool man, good luck”

(p.s. He's unemployed, I had the choice of going to many different places to work – Seattle, San Francisco, stay in my home state, work for government, etc….)

Ah….sorry. I think I just hate law students, and some lawyers (the ones that don't grow up from that law-student mentality)

REG - 92
AUD - 90
BEC - 82
FAR - 82
BISK Review Materials
DONE! /Happydance

[zzzzchen]

I am a China's CPA candidate and after reading the discussion I'd like to introduce to you CICPA, it is just insane.

It has 2 stages. In stage 1 there are 6 subjects, each has a passing rate of 20-25% percent; and each has a content about twice as much as a AICPA subject.
The exam is held once a year and most people spend 3-4 years to pass all the subjects. Studying time recommended, 1200-1500 hours.

The second stage is an all-day comprehensive test in which you get tested all 6 subjects at once, in a interconnected way. Given the participants for this test have passed all 6 subjects in stage 1, the passing rate is higher from 60-75%.

The credit requirement for enrollment is not as strict for CICPA. At the minimum a 3-year college graduate can register for the exam. But the low threshold has nothing to do with its toughness.

For myself, I passed all three CFA levels and spent about 600 hours in total. But I spent 700-800 hours preparing for just 3 subjects of CICPA.

All in all, I think it is fair to say CICPA is 2 to 3 times as demanding and tough as AICPA. Given AICPA is already hard, the CICPA is almost inhuman.

So as you can see, things are actually not that bad. Best wishes!

[Author]

Replies