For each of the LR sections, what might the general proportion of very hard and hard (4 and 5 dots on the difficulty scale) to medium and easy questions?
The test isn't curved, but it is designed to produce results resembling a bell curve. It seems like to produce this, the difficulty would need to follow that pattern. It may be a little skewed since there are other elements--like time pressure--contributing to the difficulty of the test as a whole, but it'd need to be somewhere in the neighborhood of a bell curve. I haven't gone through the data, but I'd suspect that would be the trend.
LR difficulty can be made in a variety of different manner: confusing language, useless material, multiple assumptions through 1 conclusion and 3 premises, or, my favourite-gives an circumstance and asks for the application of a principle, where you have to draw an principle, then assess the answers.
But this all comes down to a test of your tool kit as of how well do you understand what each question asks you to do and how well you are able to solve some questions with interchangeable skills.
Comments
LR difficulty can be made in a variety of different manner: confusing language, useless material, multiple assumptions through 1 conclusion and 3 premises, or, my favourite-gives an circumstance and asks for the application of a principle, where you have to draw an principle, then assess the answers.
But this all comes down to a test of your tool kit as of how well do you understand what each question asks you to do and how well you are able to solve some questions with interchangeable skills.