LSAT CURVE OPINIONS Forum

[bodadeous]

What do you guys predict the curve will be? the typical -10 = 170? A lot of people were put off by the first experimental, but maybe I am just making excuses since i was one of them and it won't lower the curve ... but i honestly think -11 to -12 will be 170

[bodadeous]

COMMON PEOPLE SACK UP AND PUT WHAT YOU THINK

[LSATnooblet]

I think it'll either be a -9 or -10

[dsin]

What does -9 -10 etc.. mean?

The sept lsat admin a 150 was 59 questions correct, and a 160 was 75 or 77 questions correct.

[zipfly]

dsin

just the number wrong for the curve....59 correct = -42/-41 etc. etc.

[wallyw]

I'm gonna go out on a limb and say -12/-13=170 because that makes me feel better!

[Ribbit]

I figure it will be a -10 = 170. We'll see. Hopefully I'm wrong.

[spursfaninla]

here are the score needed to get a 170 going backwards from dec 05:
http://www.powerscore.com/lsat/help/cor ... rgeted.htm

d 05 d 99
93, 90, 90, 87, 90, 89, 91, 88, 88, 91, 89, 89, 88, 87, 87....
http://www.blueprintprep.com/resources2.php

Above are overlayed curves for each of 2005's released LSATs and the curve for the Dec. 2002 LSAT. One can see a clear progression, especially in the middle of the curves, where more correct answers are required to attain the same scaled score.

Whereas in the December 2002 LSAT, 72 right answers were required for a score of 160, in the December 2005 LSAT, 79 right answers were required. This means that test takers had to get 7 more questions right (out of 100) on the more recent test to get the same score. That's a massive variance compared with the 1 or 2 questions that LSATs have historically varied. Bottom line: The curve is getting more difficult. You need to get more questions right than previous test-takers to get the same score.

[Eremite]

True, however (and this is going to sound crazy) I think the LSAT is getting easier. When I take those steep-curved tests from the early aughts I usually miss more. I think that the later LSATs take a more even-keeled approach and aim for the -10=170 curve.

[Fingersxed]

sad

[the lsat hax0r]

http://www.princetonreview.com/law/test ... AT-REPORTS

Based on this, I would think a -10 = 170.

[spursfaninla]

Did anyone see Kaplan rate this LSAT? I didn't see anything from them. The Princeton one saying it was slightly harder than the others from this year was VERY reassuring...

[Teflon1974]

I wouldn't place too much stock in the Princeton Review Report.

From the September 2006 Report:
As always, test forms for the September 2006 LSAT included an experimental (unscored) section. On most test forms, the experimental section appeared in Section 2. Be aware, however, that LSAC often administers a few selected forms with identical scored sections, but with their experimental sections in a different location. Ratings of this administration placed its difficulty slightly above that of other recent exams. Princeton Review students report being pleased with their overall performance.

From the December 2006 Report:
As always, test forms for the December 2006 LSAT included an experimental (unscored) section. On most test forms, the experimental section appeared in Section 1. Be aware, however, that LSAC often administers a few selected forms with identical scored sections, but with their experimental sections in a different location. Ratings of this administration placed its difficulty slightly above that of other recent exams. Princeton Review students report being pleased with their overall performance.

They played the same song for June, although they did say that February's was slightly easier. These reports are too generalized and non-specific to have much meaning.

[rsuelzer]

I thought that LSAT curve was based upon the results over the last three years. So it is possible for everyone to get a 180, but it makes the curve harder for the following year.

[Mr. Smith]

I believe that the curve is getting looser, but that's just as good of a guess as anyone's.

I think it has to do with the pool of unprepared/ill-prepared test-takers increasing exponentially in response to the recession.

[tartugas]

Why even worry about the curve? It's out of your control. If you ask me (and you sorta did by posting this q), I think it's a waste of energy to even think about the curve. Spend your time studying and aggressively attack the areas you feel weak in. If you walk in confident that you can get 10 wrong and still be ok then you're walking in with the wrong attitude.

IMHO, of course.

[tomwatts]

I thought that LSAT curve was based upon the results over the last three years. So it is possible for everyone to get a 180, but it makes the curve harder for the following year.

No, it's based on results over the last 19 years. The test is still equated by the same standards as it was back in 1991, when the current LSAT debuted, and it's never needed to be recentered, nor have scores drifted. (This is, by the way, astonishing when compared with other major standardized admissions tests.)

Basically, they take a bunch of experimental sections and slap them together to make an LSAT. Those experimental sections could be from 1992. They could be from 2009. They could be a mishmash of different years. But the scoring grid is based on historical student performance (whenever it happened) on those sections.

They do talk about percentiles from the last three years. I don't think that figures into the equating process, though.

[vivalalsat]

Why do people worry about and analyze the curve so much? You can't control it. Do something more productive please.

[kaydish21]

I thought that LSAT curve was based upon the results over the last three years. So it is possible for everyone to get a 180, but it makes the curve harder for the following year.

No, it's based on results over the last 19 years. The test is still equated by the same standards as it was back in 1991, when the current LSAT debuted, and it's never needed to be recentered, nor have scores drifted. (This is, by the way, astonishing when compared with other major standardized admissions tests.)

Basically, they take a bunch of experimental sections and slap them together to make an LSAT. Those experimental sections could be from 1992. They could be from 2009. They could be a mishmash of different years. But the scoring grid is based on historical student performance (whenever it happened) on those sections.

They do talk about percentiles from the last three years. I don't think that figures into the equating process, though.

This is the right idea, but it goes too far and is a bit too strong. They do not replicate entire experimental sections on future tests and the types of questions have evolved quite a bit since 1991. Experimental sections test specific questions on LR, and specific puzzles and passages on these respective sections. It does not make sense for LSAC to throw a whole off the wall experimental section in a test as experienced testers will quickly see this and it would be harder to interpret the results as their would be less validity in the section's scores. The way experimental works is they look at a specific question or a new twist on an existing question type and see if the test takers still get it correct at the same rate as other sections. They then have a huge database of experimentals they can use and future experimentals are used to keep up the variety of question types in this database. There is some repetition among experimentals the same way there is some repetition among non experimentals.

Sorry if this is all a bit convoluted, my head is pounding from allergies or something. Hope the brief overview helps.

[JurisDoctorate]

I believe that the curve is getting looser, but that's just as good of a guess as anyone's.

I think it has to do with the pool of unprepared/ill-prepared test-takers increasing exponentially in response to the recession.

This is my sense. I don't feel that the curve is set a priori, I feel it is made after the look at the distributions of correct answers. Since law school applications are up a bunch, and those who come late to the party are less likely to be well-prepared, the curve is getting more generous. Why else do you think the curve is getting looser and looser?

A lot of people say it's because those tests were hard; however, they are saying this after the fact. A lot of sage members are saying that people actually thought those tests were easy, at the time, and were predicting -10 and -11.

My belief is the curve will only get looser and looser; get ready for a -15 on this test.

[littlepixie11]

I believe that the curve is getting looser, but that's just as good of a guess as anyone's.

I think it has to do with the pool of unprepared/ill-prepared test-takers increasing exponentially in response to the recession.

This is my sense. I don't feel that the curve is set a priori, I feel it is made after the look at the distributions of correct answers. Since law school applications are up a bunch, and those who come late to the party are less likely to be well-prepared, the curve is getting more generous. Why else do you think the curve is getting looser and looser?

A lot of people say it's because those tests were hard; however, they are saying this after the fact. A lot of sage members are saying that people actually thought those tests were easy, at the time, and were predicting -10 and -11.

My belief is the curve will only get looser and looser; get ready for a -15 on this test.

Hm Im not too sure about that. Dec 09 gave a good curve, and June had a smaller curve. Going by that Id say its going to be -12 or -11, and probably -11. Definitely not -9 or whatever.

[DrackedaryMaster]

I was leaning at -11 yesterday but now more leaning towards -12.

Go look up the Oct 2008 TLS threads and read for yourself what those testers who had our RC as their experimental. Check the profiles/statistics of those who thought it was hard and compare to what they announced in the score release. Indications are that they struggled with it, which is surprising.

Do the same thing in the Feb 2009 threads that had our LG as experimental. They can't help us determine a curve, but many of those people were relieved it was an experimental and the "easier" LG counted.

Do the same same in the Dec 2009 in regards to the LR experimental containing the Journalist Q. Those people didn't like that LR section and it didn't make matters all that great for them considering that test as a whole was a beast.

So I'm leaning much more now that 89/90 will be 170.

[JurisDoctorate]

Here's the thing, I think the unprepared students are less likely to take the June test. They just finished their undergrad, they are still on their "well-deserved" vacations and they don't start thinking about their future until the fall.

Mark my words, the curve is going to be huge. And, as one poster dreaded, it'll be -20 in no time. However, you'll then need like a minimum of 175 to get into CCN.

Times are a changin'