How are questions ranked in difficulty

MIT_2017

I notice when I'm reviewing there is a difficulty rating both for individual questions as well as overall sections. Is this done by JY or by users? Did I miss somewhere in the CC where this is explained (admittedly, I came to 7Sage after already scoring low 170s so I did not look at most of the CC...)

drbrown2

It’s based on 7Sage user data for those questions/sections

MIT_2017

@drbrown2 said:
It’s based on 7Sage user data for those questions/sections

So that means how users have scored on the questions, right? And not them self-reporting the difficulty.

drbrown2

@MIT_2017 said:

@drbrown2 said:
It’s based on 7Sage user data for those questions/sections

So that means how users have scored on the questions, right? And not them self-reporting the difficulty.

Correct. Not sure how big the pool of data is from over the years but I assume there are less data for newer tests. There are varying levels of difficulty for different people. We all have different ways of thinking and naturally make different assumptions while we read, so some 5 star questions feel a little bit easier for different people. Hard questions are hard but obviously when you are reading and you clue in on the key to the question it is much easier to get it right under timed conditions. The majority of test takers aren't picking up on those during timed conditions. Theoretically a student who blindly guesses on any questions would have 20% chance of getting it right (if they truly chose between A and E at random), but I wonder if the odds of getting the correct answer on the very difficult questions are lower than 20% due to trap answer choices and subtlety.