better understanding my mean score

Lear22
Posts: 275
Joined: Wed Sep 28, 2011 10:17 am

better understanding my mean score

Postby Lear22 » Wed Sep 26, 2012 8:51 pm

As I move to all PTs until my Oct 10th test, I am trying to get a better analysis of the mean score, since the actual score according to the mean score changes from test to test.

It appears that in some tests for example, you had to score an 80 to get a 165, while in others you needed as much as 85 (pretty big difference). Is that because that specific test was harder, or because more people sat in that test date? Do you guys measure your mean scores when comparing between different PTs that you take or the actual score? (since again, you could score the same amount of points in 3 PTs, but the actual score will change).

It also appears that the mean score is going down on recent exams compared to earlier years. From the looks of it it seem that tests in '08 through '09 (if looking at more recent test dates) asked for a higher mean for a better score (if we look at 165 again for example, most of them were 83, 84 and 85).

Thanks for any help with this. Just trying to get clarity and understanding of how I preform and how will that resonate on my actual LSAT.

PS: As I understand it, non-saturday test takers (as I am one) are put in the same pool as the rest of the test takers, correct?

User avatar
nick_scheu
Posts: 64
Joined: Mon Sep 17, 2012 3:32 pm

Re: better understanding my mean score

Postby nick_scheu » Wed Sep 26, 2012 9:07 pm

The raw score (typically out of 101, but it varies) is, as you know, the actual number of questions you got right. The scaled score (120-180) takes the raw score and converts it to a form that can be compared between LSATs. That is, in theory, a person who got a 170 on PT45 would have received the exact same score on PT65 if he had taken it instead, within a margin-of-error (2-3 points, IIRC).

The conversion from raw to scaled is not actually a curve. A curve is calculated based on the distribution of scores for a particular administration, meaning that if only really smart people took the test, you'd get nailed (and you'd have a huge benefit if you were the only smart guy and everyone else was exceptionally dumb).

The scaled score doesn't work that way. Instead, it uses data from the experimental sections administered in prior LSATs to determine the actual difficulty of a particular test. The performance of others is irrelevant. Easier LSATs allow you to miss fewer raw points; harder LSATs allow you to miss more.

In other words, any score you get on a PT should be a true, real score for that particular moment in time. If you're getting wildly inconsistent scores, it's likely lucky guessing. The scaling is not based on the performance of the students who took that LSAT for real.

User avatar
northwood
Posts: 4872
Joined: Fri May 14, 2010 7:29 pm

Re: better understanding my mean score

Postby northwood » Wed Sep 26, 2012 9:10 pm

Generally, the more difficult the test, the lower your raw score is to get a higher LSAT number. I suck at math, so im going to go with a goal of 170. If the test is hard then lets say you need to get 87 out of 100 to get a 170( some tests have 101 questions, but to make math easy- im only going to go with 100- because the concept is the same- but with 1 extra question) . That means you can miss 13 questions and end up with a 170. This also meansthe curve is -13. If the test was considered easier then lets say you can only miss 9 questions to get teh 170. That would mean you have to answer 91 questions correctly to score the same LSAT score. ANd thus the curve is -9. SO to get the same score, on an " easier" test you need to answer more questions correctly than on harder tests.

This is also why your scoreon the actual thing is you score +/- 3. That means, that if you were to take another test you should expect to score within 3 points above or below that test days score ( which is why a 3 point swing + or - is not a big jump in most schools opinion).

Dont worry about the real LSAT score, focus on your raw ( out of 100) score. You should be tracking the number of questions you got right ( and for each section) and seeing how you do. SO for example you average a score of 93- then you know if the curve is -7 you should get a 170... but if the curve is -8 or -9, -11, etc you will score better, even thouuh you answered the same number of questions correct. Since there is no way you can predict the curve, or should even bother trying to- this is a good way to gague where you should expect to be when you get your score back ( and if you should postpone it or keep it- postpone if you are not happy with your raw score and its coorelating average real score keepit if you are) which may tell you if you should retake or not. hope this helps

Lear22
Posts: 275
Joined: Wed Sep 28, 2011 10:17 am

Re: better understanding my mean score

Postby Lear22 » Wed Sep 26, 2012 10:22 pm

northwood wrote:Generally, the more difficult the test, the lower your raw score is to get a higher LSAT number. I suck at math, so im going to go with a goal of 170. If the test is hard then lets say you need to get 87 out of 100 to get a 170( some tests have 101 questions, but to make math easy- im only going to go with 100- because the concept is the same- but with 1 extra question) . That means you can miss 13 questions and end up with a 170. This also meansthe curve is -13. If the test was considered easier then lets say you can only miss 9 questions to get teh 170. That would mean you have to answer 91 questions correctly to score the same LSAT score. ANd thus the curve is -9. SO to get the same score, on an " easier" test you need to answer more questions correctly than on harder tests.

This is also why your scoreon the actual thing is you score +/- 3. That means, that if you were to take another test you should expect to score within 3 points above or below that test days score ( which is why a 3 point swing + or - is not a big jump in most schools opinion).

Dont worry about the real LSAT score, focus on your raw ( out of 100) score. You should be tracking the number of questions you got right ( and for each section) and seeing how you do. SO for example you average a score of 93- then you know if the curve is -7 you should get a 170... but if the curve is -8 or -9, -11, etc you will score better, even thouuh you answered the same number of questions correct. Since there is no way you can predict the curve, or should even bother trying to- this is a good way to gague where you should expect to be when you get your score back ( and if you should postpone it or keep it- postpone if you are not happy with your raw score and its coorelating average real score keepit if you are) which may tell you if you should retake or not. hope this helps

ֿ
It does! thanks!!

How about the curve for the non-sat LSAT? Anyone knows?
thanks again




Return to “LSAT Prep and Discussion Forum”

Who is online

Users browsing this forum: Yahoo [Bot] and 2 guests