[slickrick]

Has AICPA ever said why a candidate doesn’t receive their scores right after testing?

[Recked]

It's 100% not because there is a curve. I repeat NOT because there is a curve.
I think the official explanation is so that they can “statistically validate the scores”. But its not a curve.
The questions are assigned a certain number of points based on difficulty, which is determined presumably by how many people get it correct or incorrect.
I imagine if too many people fail the test they can adjust the individual weights of the questions to meet desired statistical goals.
But what do I know…

Memento Mori - Kingston NY CPA & EA (SUNY Albany 2002)

FAR-93 11/9/17 (10wks, 250 hrs, Roger 1800+ MCQs, Gleim TB 600+MCQs, SIMs)
AUD-88 12/7/17 (3 wks, 85 hrs, Roger 1000 MCQs no SIMs hail mary)
REG-96 1/18/18 (6 wks, 110 hrs, 1400 MCQs, no SIMs)
BEC-91 2/16/18 (4wks, 90 hrs, 1240 MCQs)

[DDylan]

“The questions are assigned a certain number of points based on difficulty, which is determined presumably by how many people get it correct or incorrect.
I imagine if too many people fail the test they can adjust the individual weights of the questions to meet desired statistical goals.”

—the process you described sounds like a curve tho

[slickrick]

THIS LEAVES ONE TO BELEIVE THAT SOMETHING FISHY IS GOING ON BEHIND CLOSED DOORS. @ RECKED…SO BASICALLY IF EVERYONE TANKED THE TEST, THEN THE AICPA WOULD HAVE TO ADJUST EVERYONE UP TO PASSING, RIGHT?

[Recked]

sense the sarcasm in my post?
if everyone tanked, approximately 40-50% would probably still pass, LOL

Memento Mori - Kingston NY CPA & EA (SUNY Albany 2002)

FAR-93 11/9/17 (10wks, 250 hrs, Roger 1800+ MCQs, Gleim TB 600+MCQs, SIMs)
AUD-88 12/7/17 (3 wks, 85 hrs, Roger 1000 MCQs no SIMs hail mary)
REG-96 1/18/18 (6 wks, 110 hrs, 1400 MCQs, no SIMs)
BEC-91 2/16/18 (4wks, 90 hrs, 1240 MCQs)

[turo9992000]

They say that the test is always the same difficulty and that if some quarters have a higher pass rate it is because those candidates were better prepared.

AUD - 64, 80 Passed on 10/09/17

BEC - 75 Passed on 12/09/17

FAR - 69, 71, 73, 83 Passed on 06/10/18

REG - 81 Passed on 09/10/17

[RoseMarie]

One reason is because Prometric still has to submit the tests to the AICPA which can take up to 24 hours according to the score release pages notes. Also while they may not call it a curve, I personally believe if everyone was failing, especially after major changes like the recent REG changes, then they would do some adjusting of weights or throw out questions that most people missed or something.

FAR - 11/05/18 - 92

BEC - 12/04/18 - 95

AUD - 2/27/19 - 96

REG - 4/17/19 - 93

I’m done!!

 

[DM]

There is a major historical component to the exam. IMO, old questions are recycled to serve as a baseline for each exam and these questions are expected to be answered correctly for a CPA candidate. If everyone fails, then everyone fails. The older questions are “locked” with point value; more point value in the harder testlets. There is no curve. Once a new question is tested over and over, and reaches a statistical measure the AICPA is happy with then that question will be expected to be answered correctly by CPA candidates.
I think the test grading goes like this; everyone starts at a “75”. If you ace the medium testlet, then you have a chance to improve on the “75” in the harder testlet. You can still hurt yourself in the harder testlets, but not as bad as the medium testlets. If you get medium testlet questions wrong, then you are below the pass threshold. IMO, the short answer is heavily weighted based on that testing period's responses. If you stay in the “medium testlets” then your short answer is very important to pass and may be the difference. I think the harder testlets are worth much more in point value and short answer is not that important to overall score. There is a reason that short answer is at the end of the exam. I think MCQ is make or break.

[DM]

CPA candidates are tested against the material, not each other. The AICPA regulates the questions on the exams to what is expected to be known by the CPA candidate. So you are going against a benchmark, not the other test takers. The AIPCA can control the pass rates by regulating the benchmark.

[slickrick]

@ DM…..You present an interesting theory on the scoring process.

[Brokenglass]

I got all of this info from the rogercpareview website. So first know that there are pretest questions on each of the four sections, which include 12 pretest MCQs and 1 pretest simulation. For BEC, one of the written communication is also a pretest. So for all sections 12 MCQ and 1 simulation (and 1 WC for BEC) aren't considered in grading.
That alone explains why you never know how well you did on an exam.
The first testlet is always medium difficulty. If you do well, the second testlet will be hard difficulty. If you do poorly, the second teslet will stay at medium difficulty. This only applies to MCQ. More weight is assigned to harder MC questions, so you get more points for each hard question you get right.
For BEC, the WC are graded by a computer program. If it determines that you passed, then you are given a specific score based on additional review by an actual person.

AUD 77 (6/2018)

REG 77 (9/2018)

FAR 61 (12/2018), 84 (1/2019)

BEC 88 (2/2019)

Studying time using Becker (4-6 hours)

AUD: 16 days

REG: 13 days (skipped units 7 and 8 and did only flashcards instead)

FAR: 28 days, 12 days, skipped lectures for both attempts

BEC 13 days

[Steven K]

There is a lot of conspiracy that the AICPA does this in order to properly weigh the questions to achieve pass rates around the 50% range. However, I am not sure if I agree with this. One thing I will say, and I'm not sure if this is still relevant, is because they have to manually evaluate scores that are close.
I had a professor in undergrad who used to grade the CPA exam back when it was still on paper. He said that although most people believe the AICPA wants you to fail, this is actually not the case. He told my class that if you got a 75 on the exam, you probably actually scored around a 71 or 72. This is because if you were that close to passing, the AICPA would manually review your score and try to give you as many points as possible for the SIMS. Again, not sure if this has drastically changed since the exam is now done online however, this advice did come from someone who used to sit behind those closed doors.

"If you put in the extra effort at the beginning, the process won't take as long."

FAR - 81 (11/10/18)
AUD - 79 (12/8/18)
REG - 84 (1/22/19)
BEC - 84 (2/13/2019)

[jombe]

1. It's pretty shady how the exams are graded, since no one can explain black & white how they are graded.
2. It's the freaking gov't. It's like most gov't entities use technologies from a decade or two ago. I just got exposed to the world of sales tax at my new position and where I file my company's sales tax is a joke. The interface looks soooo old. A few jurisdictions even require me to submit IN PAPER. It's one of those thin carbon copy ones. Lol.

AUD - 99
BEC - 91
FAR - 94
REG - 96

--------------------------------------------
Done with exam. On with life.

FAR - 94 (10/4/15), Local Prep Online Lectures, BISK & NINJA MCQ
AUD - 99 (1/19/16), Local Prep Online Lectures, BISK & NINJA MCQ
REG - 96 (4/19/16), Local Prep Online Lectures, BISK & NINJA MCQ
BEC - 91 (7/19/16), Local Prep Online Lectures, BISK & NINJA MCQ

581 days of listening to lectures, reading texts & 10,000+ MCQs...

[12tang]

I firmly believe one of the major reasons why they don't release scores sooner is because it's one more aspect to the game of obtaining a CPA licensure that keeps candidates crapping in their pants… They love to keep candidates sweating, miserable and hungry for more abuse…

BEC - PASS

FAR - PASS

AUD - PASS

REG - PASS

BOOM!  JUST LIKE THAT, I GOT MY LIFE BACK!  =D

Using Becker self-study
FAR: (82) 175 hours - 1st attempt
BEC: (XX)
AUD: (69) 45hrs of study - 1st attempt
REG: (XX)

[Author]

Replies