LSAT score and law school performance Forum

[Alex-Trof]

A Virginia law dean mentioned the LSAT correlation as about 0.40 when used to predict performance during the first year of law school, and about a 0.50 correlation when paired with one's undergraduate GPA. This was a videotaped lecture regarding whether the LSAT should be optional (the presenter was a former ABA education chairman) given to AA students attending or accepted to Virginia law.

http://www.lsac.org/LsacResources/Resea ... orrelation

.33

P.S. There was already a thread on this.

http://www.top-law-schools.com/forums/v ... n&start=25

[dreadlawks]

Minute consideration:

Can "most reliable" be interpreted as the closest thing we have to reliable without actually being reliable?

[bk1]

Minute consideration:

Can "most reliable" be interpreted as the closest thing we have to reliable without actually being reliable?

In laymen's terms I guess you could. But considering reliability isn't binary it's hard to make that argument unless you are setting a firm cutoff for reliability (though there are cutoffs for certain definitions of reliability in certain instances).

The main thing is that it has been shown to be pretty reliable when it is studied.

[09042014]

Minute consideration:

Can "most reliable" be interpreted as the closest thing we have to reliable without actually being reliable?

The more important question is what else are you going to use? Every argument against LSAT is even stronger against GPA. It's less reliable a predictor. It's biased against minorities and the poor. It's not standardized at all (3.6 MIT EE being the same as 3.6 criminal justice from Western Illinois Uni? LOL).

Go back to the system where Harvard Undergrads got into HLS, and UIUC grads couldn't? Not reliable, discriminatory and you are just replacing an LSAT from 9 months ago with a SAT from 6 years ago.

Law school admissions would turn into who can best bullshit a personal statement (or worse, who can buy the best personal statement).

Getting rid of the single best indicator of performance because it doesn't perfectly predict actual grades is silly unless you come up with a better indicator first.

[pinkzeppelin]

Minute consideration:

Can "most reliable" be interpreted as the closest thing we have to reliable without actually being reliable?

The more important question is what else are you going to use? Every argument against LSAT is even stronger against GPA. It's less reliable a predictor. It's biased against minorities and the poor. It's not standardized at all (3.6 MIT EE being the same as 3.6 criminal justice from Western Illinois Uni? LOL).

Go back to the system where Harvard Undergrads got into HLS, and UIUC grads couldn't? Not reliable, discriminatory and you are just replacing an LSAT from 9 months ago with a SAT from 6 years ago.

Law school admissions would turn into who can best bullshit a personal statement (or worse, who can buy the best personal statement).

Getting rid of the single best indicator of performance because it doesn't perfectly predict actual grades is silly unless you come up with a better indicator first.

TITCR!

And taxguy, quit trolling against the LSAT. Telling a bunch of people online that the LSAT is a bad predictor of success will not magically make your son outperform his abysmal LSAT score.

[bk1]

Taxguy always manages to disappear when his bullshit statements get refuted.

Makes me sad.

[09042014]

I wondered why a supposedly old lawyer would troll against the LSAT so hard. It makes sense it's because of his retarded son.

[bk1]

I wondered why a supposedly old lawyer would troll against the LSAT so hard. It makes sense it's because of his retarded son.

His son is a special snowflake that will succeed no matter what!

[taxguy]

Gee why would I rail against the LSAT? IT has NOTHING to do with my son. I have met many, many lawyers who were outliers. It is just too many of them. I have also met many lawyers being a continuing education provider and found many of them become outliers from the LSAT.

Admittedly, I didn't do a statistical study. Admittedly, I didn't meet every lawyer in America or even get a necessarily random sample. However, I have met too many lawyers whose performance didn't corelate to the LSAT and have also met many who did. It almost seems like a crapshoot.For what it's worth, I have met a number of law school deans and professors and most of them admit that there have been a large number of outliers from what the LSAT would have predicted. Yes, this isn't a studied, random sample.

[09042014]

Gee why would I rail against the LSAT? IT has NOTHING to do with my son. I have met many, many lawyers who were outliers. It is just too many of them. I have also met many lawyers being a continuing education provider and found many of them become outliers from the LSAT.

Admittedly, I didn't do a statistical study. Admittedly, I didn't meet every lawyer in America or even get a necessarily random sample. However, I have met too many lawyers whose performance didn't corelate to the LSAT and have also met many who did. It almost seems like a crapshoot.For what it's worth, I have met a number of law school deans and professors and most of them admit that there have been a large number of outliers from what the LSAT would have predicted. Yes, this isn't a studied, random sample.

And yet a studied random sample shows a fair amount of correlation, even when classes are already sorted by LSAT. It is the best way to pick a class.

I'll even admit that schools improperly use the LSAT. They use it like it is extremely precise, when it's not.

What do you think they should use instead?

[bk1]

Gee why would I rail against the LSAT? IT has NOTHING to do with my son. I have met many, many lawyers who were outliers. It is just too many of them. I have also met many lawyers being a continuing education provider and found many of them become outliers from the LSAT.

Admittedly, I didn't do a statistical study. Admittedly, I didn't meet every lawyer in America or even get a necessarily random sample. However, I have met too many lawyers whose performance didn't corelate to the LSAT and have also met many who did. It almost seems like a crapshoot.For what it's worth, I have met a number of law school deans and professors and most of them admit that there have been a large number of outliers from what the LSAT would have predicted. Yes, this isn't a studied, random sample.

When there are tens of thousands of JDs awarded every year (and all of them took the LSAT), the sheer number of outliers is going to be huge.

Again, the fact that you rail against something when the statistics go against your random anecdotes is telling of your understand of math. It's not surprising that they say people go to law school because they can't do math.

[09042014]

Gee why would I rail against the LSAT? IT has NOTHING to do with my son. I have met many, many lawyers who were outliers. It is just too many of them. I have also met many lawyers being a continuing education provider and found many of them become outliers from the LSAT.

Admittedly, I didn't do a statistical study. Admittedly, I didn't meet every lawyer in America or even get a necessarily random sample. However, I have met too many lawyers whose performance didn't corelate to the LSAT and have also met many who did. It almost seems like a crapshoot.For what it's worth, I have met a number of law school deans and professors and most of them admit that there have been a large number of outliers from what the LSAT would have predicted. Yes, this isn't a studied, random sample.

When there are tens of thousands of JDs awarded every year (and all of them took the LSAT), the sheer number of outliers is going to be huge.

Again, the fact that you rail against something when the statistics go against your random anecdotes is telling of your understand of math. It's not surprising that they say people go to law school because they can't do math.

Also LSAT doesn't attempt to predict success as a lawyer. It's a measurement to predict success as a first year law student.

[BarbellDreams]

I dont really have an idea of how to make the LSAT better at predicting law school success but I think that since this is TLS people with high LSATs will troll that the LSAT is the best predictor ever and people with low LSATs will say it makes no difference for law school.

Speaking from strictly personal experience, I had a very average LSAT score (at least in my view) and due to unique circumstances did not retake and went to law school instead. My 1L grades were much better than anything I ever expected and I know plenty of people who got up to 8 points higher than had lower grades. FWIW.

[09042014]

I dont really have an idea of how to make the LSAT better at predicting law school success but I think that since this is TLS people with high LSATs will troll that the LSAT is the best predictor ever and people with low LSATs will say it makes no difference for law school.

Speaking from strictly personal experience, I had a very average LSAT score (at least in my view) and due to unique circumstances did not retake and went to law school instead. My 1L grades were much better than anything I ever expected and I know plenty of people who got up to 8 points higher than had lower grades. FWIW.

Learn about statistics please.

[bk1]

I dont really have an idea of how to make the LSAT better at predicting law school success but I think that since this is TLS people with high LSATs will troll that the LSAT is the best predictor ever and people with low LSATs will say it makes no difference for law school.

Speaking from strictly personal experience, I had a very average LSAT score (at least in my view) and due to unique circumstances did not retake and went to law school instead. My 1L grades were much better than anything I ever expected and I know plenty of people who got up to 8 points higher than had lower grades. FWIW.

[BarbellDreams]

I dont really have an idea of how to make the LSAT better at predicting law school success but I think that since this is TLS people with high LSATs will troll that the LSAT is the best predictor ever and people with low LSATs will say it makes no difference for law school.

Speaking from strictly personal experience, I had a very average LSAT score (at least in my view) and due to unique circumstances did not retake and went to law school instead. My 1L grades were much better than anything I ever expected and I know plenty of people who got up to 8 points higher than had lower grades. FWIW.

Thanks for that.

Anyway, I am not hating on the LSAT. And I am aware that there are some statistics saying there is a correlation. It just seems to me that these statistics aren't as representative as some would have them be. At least at my school I am far from an anomaly in terms of mediocre-LSAT-to-good-grades.

[bk1]

Anyway, I am not hating on the LSAT. And I am aware that there are some statistics saying there is a correlation. It just seems to me that these statistics aren't as representative as some would have them be. At least at my school I am far from an anomaly in terms of mediocre-LSAT-to-good-grades.

Trust me, you are an outlier. Hope this link helps explain that to you: http://en.wikipedia.org/wiki/Outlier

[Alex-Trof]

It is interesting that LSAT/L1 grades correlation is decreasing while undergrad GPA/L1 grades correlation is increasing. LSAC is not really giving any explanation to why it is happening.

[09042014]

It is interesting that LSAT/L1 grades correlation is decreasing while undergrad GPA/L1 grades correlation is increasing. LSAC is not really giving any explanation to why it is happening.

Law school classes are ever more tightly grouped by LSAT. If everyone had a 170 at a school the correlation would be 0.

[Alex-Trof]

It is interesting that LSAT/L1 grades correlation is decreasing while undergrad GPA/L1 grades correlation is increasing. LSAC is not really giving any explanation to why it is happening.

Law school classes are ever more tightly grouped by LSAT. If everyone had a 170 at a school the correlation would be 0.

Makes sense. But what about GPA. Didn't it go from .15 to .28?

[09042014]

It is interesting that LSAT/L1 grades correlation is decreasing while undergrad GPA/L1 grades correlation is increasing. LSAC is not really giving any explanation to why it is happening.

Law school classes are ever more tightly grouped by LSAT. If everyone had a 170 at a school the correlation would be 0.

Makes sense. But what about GPA. Didn't it go from .15 to .28?

I don't think it was ever .15.

I thought it used to be .4 LSAT .3 GPA.

[JusticeHarlan]

Gee why would I rail against the LSAT? IT has NOTHING to do with my son. I have met many, many lawyers who were outliers. It is just too many of them. I have also met many lawyers being a continuing education provider and found many of them become outliers from the LSAT.

Admittedly, I didn't do a statistical study. Admittedly, I didn't meet every lawyer in America or even get a necessarily random sample. However, I have met too many lawyers whose performance didn't corelate to the LSAT and have also met many who did. It almost seems like a crapshoot.For what it's worth, I have met a number of law school deans and professors and most of them admit that there have been a large number of outliers from what the LSAT would have predicted. Yes, this isn't a studied, random sample.

And yet a studied random sample shows a fair amount of correlation, even when classes are already sorted by LSAT. It is the best way to pick a class.

I'll even admit that schools improperly use the LSAT. They use it like it is extremely precise, when it's not.

What do you think they should use instead?

He thinks we should look at if you've written chapters of an as-yet incomplete book, started a fraternity, and played ultimate frisbee.