Law School Discussion

Nine Years of Discussion
;

Author Topic: Is there a correlation between LSAT Score and First Year Performance?  (Read 3982 times)

masterthegame

  • Full Member
  • ***
  • Posts: 16
    • View Profile
    • Email
MY Testmaster teacher told us there is a strong correlation between the LSAT and first year performance in law school. He said "90 % of students who do really well on the LSAT also do well in first year law school. It's been proven. If you have a tough time on this test you are likely to struggle in law school."

Is this true? I'm not doing "really well". Should I give up my hope for law school since everything pretty much matters on your first year.

Anyone know of this study or could cite it. I couldn't find it on the internet.

Bulldog86

  • Sr. Citizen
  • ****
  • Posts: 207
    • View Profile
Google LSAT correlation study - the first result (for me) is a recent one from 2003-2004. (Direct PDF link)

The short version: There's a pretty strong correlation between LSAT and first-year grades; a better correlation between (LSAT + GPA) and first-year grades; and a weaker correlation between GPA alone and first-year grades.

LSAC spends a lot of effort making sure this correlation exists; if it didn't, nobody would care about the LSAT.

I wouldn't take your initial struggles as a sign you should quit though, and that's pretty crappy of the teacher to say that. Fact is, the scores used in this correlation study are largely scores that were received after significant prep work. They don't do a correlation of 1L grades with people taking their first diagnostic test cold. (Though it would be interesting!) So high LSAT scores for many people also indicate hard work -- also a good skill for law school. Whether that is the aim of the test, of course, is irrelevant.
UVA Law Class of 2011

jack24

  • Sr. Citizen
  • ****
  • Posts: 1050
    • View Profile
I've seen threads that have a lot of data on this subject but I don't have it.
I think it's very hard to measure.   How do you define "really well"
I think everyone who goes to Yale does "really well" on the LSAT but not all of them can be in the top of their class.
There are some people with 177 LSAT scores who are in the bottom half of their class.  To get a proper correlation you would have to compare LSAT scores and grades at each individual school.   Does a 175 have a statistically lower chance of doing well at Yale than a 178?   Does a 162 have a statisically better chance of doing well at Gonzaga than a 152?



LittleRussianPrincess, Esq.

  • Sr. Citizen
  • ****
  • Posts: 2355
  • Hopelessly devoted...to the Tennessee Vols!
    • View Profile
Google LSAT correlation study - the first result (for me) is a recent one from 2003-2004. (Direct PDF link to this particular one: www.lsacnet.org/research/Predictive-Validity-LSAT-National-Summary-2003-2004-Correlation-Studies.pdf)

The short version: There's a pretty strong correlation between LSAT and first-year grades; a better correlation between (LSAT + GPA) and first-year grades; and almost no correlation between GPA alone and first-year grades.

LSAC spends a lot of effort making sure this correlation exists; if it didn't, nobody would care about the LSAT.

I wouldn't take your initial struggles as a sign you should quit though, and that's pretty crappy of the teacher to say that. Fact is, the scores used in this correlation study are largely scores that were received after significant prep work. They don't do a correlation of 1L grades with people taking their first diagnostic test cold. (Though it would be interesting!) So high LSAT scores for many people also indicate hard work -- also a good skill for law school. Whether that is the aim of the test, of course, is irrelevant.

PDF's crash my comp, but isn't the figure something like the following:

LSAT + GPA predict about 25% of 1L grades?

Russian by birth, Southern by the grace of God.

Bulldog86

  • Sr. Citizen
  • ****
  • Posts: 207
    • View Profile
I'm not terribly proficient in stats, so I'm not positive these are the numbers you want, but the data summary at the end says this:

Quote from: LSAC
The average multiple correlation between first-year grades in law school and the combined predictors of LSAT and UGPA is .48 for 2003 and .47 for 2004.
...
The median validity for LSAT alone is .37 for 2003 and .35 for 2004, compared with .29 for UGPA alone for both years.

To answer Jack24's question, LSAC runs the correlation on each individual school and then looks at means, medians, Q1, Q3, etc., in the set of the combined results of the schools. You really have to, due to curves -- obviously, comparing a 175 who has a 3.1 at Harvard and a 149 who has a 3.6 at Cooley is not an apples-to-apples comparison.
UVA Law Class of 2011

The Artist

  • Sr. Citizen
  • ****
  • Posts: 321
  • Admit it... you got Rick Rolled.
    • View Profile
There's a correlation, but it's not predictive, if that makes sense. People who did well on the LSAT likely worked hard. People who do well in law school likely work hard. But there are also geniuses who never try hard and will not do well despite a good score, and there are also overachievers without raw intelligence (usually those with high GPA/low lsat) who will work their way to the top of the class. Hence why this board always implores people choosing between a school and a slightly lower ranked one with $$$ to go to the better school, because it isn't safe to assume your class rank will be higher at the lower ranked school, even if your LSAT is above 75%ile.

LittleRussianPrincess, Esq.

  • Sr. Citizen
  • ****
  • Posts: 2355
  • Hopelessly devoted...to the Tennessee Vols!
    • View Profile
Honestly, I can't understand why the correlation is taken so seriously.  Law School Exams require you to write a three hour paper about a hypothetical legal scenario -- why wouldn't a stronger correlation exist between that and the forgotten LSAT writing sample?  Why can't we grade your ability to write well and analyze like a lawyer?  Oh, yeah.  It would take too much money and effort.

Moreover, don't bother with these correlations.  You'll drive yourself crazy, or, worse, create self-fulfilling prophecies about where you'll end up.


The LSAC is actually funding ongoing research at Boalt which seeks to amend the LSAT to test more "lawyering skills." There's a pilot test around and some of us were asked to take it while I was in law school, though I didn't personally participate.
Russian by birth, Southern by the grace of God.

EarlCat

  • Global Moderator
  • Sr. Citizen
  • ****
  • Posts: 2533
  • i'm in ur LSAT blowin' ur curve
    • AOL Instant Messenger - EarlCat78
    • View Profile
    • EarlDoesLSAT.com
The listening comprehension idea was a neat one.  They actually had many people in rooms wearing headphones listening to recorded dialog that you could not ask to be repeated or whatever that they were asked to answer questions about on paper.

I think I just threw up a little.

jeffislouie

  • Sr. Citizen
  • ****
  • Posts: 413
    • View Profile
    • Email
I find this idea to be a little on the silly side, especially since the OP said that a person with a vested interest in you taking their program said it.  Besides, it is more than a little stupid considering how many people take LSAT prep courses to improve their lsat score.  Simply put: if the LSAT was a true correlary to success rate at law school, the ABA would likely discourage people from taking LSAT courses and work their hardest to keep test information outside of the public view.  Why?  Well, it's sort of like cheating to take a prep class if they truly believe an LSAT score is predictive of law school success.
For instance, I know a few people who took practice tests and scored 155, then took Kaplan and took the actual LSAT and got a 166.  What this means is that the predictive score was a 155, not the 166 they ended up with.  As a result of these courses teaching you how to take the test, the numbers are skewed and somewhat meaningless.
Add to that the fact that I know law students with LSAT scores all over the place - some with higher scores did poorly and some with lower scores made dean's list.
Then there is the group of people who simply don't test well.  I know a guy who got a 125 on his practice test and a 150 on his actual test who is in the top 10% of his class and has been for 2 straight years.
The LSAT is merely another number for law schools to look at that they feel has some impact on how well a student will do.  I'm sure someone who got a 170 on the LSAT MAY have a better chance in law school, but I've met too many people with lower LSAT scores who have done extremely well.
That said, don't start believing that if your LSAT score isn't perfect you won't do well.  It is a measuring device (in my opinion, it measures desire more than anything else) that is just as flawed as considering the color of a persons eyes or their height in deciding whether to accept or not.
Do the best you can and force yourself not to worry.  No one gives a crap what you scored on your LSAT once you get into school.
Justice is tangy....

PSUDSL08

  • Sr. Citizen
  • ****
  • Posts: 408
    • View Profile
    • LS #'s
I'm walking proof countering the correlation. I had a 151 LSAT, which put me at the bottom 25th percentile for entering students at my T4, and I'm guessing I probably have the lowest LSAT score of anyone in my class at my current T2. Scored in the top 30% at the T4, and about the top 10-12% at the T2. Probably did more work/studying than the average student, but not even close to being buried in the library like a good chunk of students were. According to the correlation, I should have been in the bottom quarter or so at my T4 and the village idiot at my T2.

I think the premise of the LSAT's importance is as follows: you can be smart and have the ability to understand, apply, and practice law and do poorly on the LSAT, but you can't be an idiot and do well on the LSAT. I guess law schools take the approach that the former need to prove they're "smart" and the latter are entitled to a presumption of intelligence as it pertains to the study of law. I don't know much about statistics, but I don't understand why a general IQ test wouldn't provide a similar correlation.

I wouldn't let a low LSAT score relative to your peers discourage you in terms of how you perform relative to your classmates.