There are a lot of good theories of how this test is scored on here. I seem to lean towards the one that means the top 50%ish during that testing window pass no matter what percentage they actually got correct. As an example:
Testing window #1. You felt awesome on the exam and got 80% of the questions right, however your 80% right is far worse than how all the other test takers did and as a result 51% of all scorers were better than 80%. You fail. The top 50% are deemed to have passed as they were the top percentage.
Testing window #2. Brutal questions, real tricky and you struggled. Everyone struggled also. You only got 65% of the test right. However, your 65% is better than half the test takers so you are in the upper 49% and you get a pass.
To me this is logical. When people get an “easy” test, many other students might get it as well so it is very hard to differentiate yourself unless you virtually get every single question right.
If it is a hard test and others are getting a really hard test also, it doesn't take much for you to get some differentiation because the top score might be low also instead of perfects in the other example.
Again this only follows suit if they are simply taking the 50% or so and saying they pass each time. Which based on results appears they might have a go / no-go line in place for passing.
This is very similar to an employee review concept called forced rankings(which my company does). We have four rankings 1-4 with 1 being the best. 1's can only be 20% of the population, 2's are 40%, 3's are 30% and 4's are 10%. Now during this process someone ranked a 3 is considered a really good employee that gets their job done successfully and completely all the time. To me making a 74 on the CPA exam equates to being a 3 in the forced rankings model. You are smart, you do a good job, however there are 60% more people out there that are doing an even better job than you, which is a little higher than actual passing rates.
To show this more clearly, assume we have 10 test takers and these are their scores:
99
92
88
87
85
83
80
77
74
65
A lot of these people did well on the test right? But forcing the rankings only means the first 6 actually pass and the bottom four do not based on their peers performance.
Doing this in a way ensures only the best people are passing each window as they might be scoring correctly much higher than 75% of the test while getting an advisory score of 75.
Or the AICPA might just throw darts and pass out scores.. who knows…
First go at the CPA! Only using Becker
Reg / Nov 2015 - 87
Far / Apr 2016 - 79
Bec / May 2016 - 80
Aud / Aug 2016
.