GU and NU accept GRE - The Barbarians at the Gate for LSAC

(Applications Advice, Letters of Recommendation . . . )

sparkytrainer
Posts: 228
Joined: Sat Jan 17, 2015 12:32 am

Re: GU and NU accept GRE - The Barbarians at the Gate for LSAC

Postby sparkytrainer » Tue Aug 08, 2017 4:29 pm



Paywall, can you post the text? But I can figure out why the lsat is scared- Gre is given every weekend, its instant scored, easier, and broader so you could apply to law school and phd programs with one test. If I was the lsat, I would be scared shitless.

azaleay
Posts: 3
Joined: Thu Feb 09, 2017 11:47 am

Re: GU and NU accept GRE - The Barbarians at the Gate for LSAC

Postby azaleay » Tue Aug 08, 2017 5:16 pm

I think to get to the same percentile, taking GRE is probably easier. GRE scores are not that important for graduate school application, as long as you meet the minimum score. At least this is true for STEM majors. When I applied for graduate school, as a STEM major, I did not study much because I did not need a super high score. So if you take GRE, you are competing with many people who do not study very hard; but for LSAT, every point matters. However, I guess schools will take this into account, and a near-perfect GRE score also needs a lot of hard work.

User avatar
TheKingLives
Posts: 66
Joined: Sun Mar 01, 2015 8:37 am

Re: GU and NU accept GRE - The Barbarians at the Gate for LSAC

Postby TheKingLives » Tue Aug 08, 2017 9:57 pm

I was able to access Gordon_Cole's link by googling "LSAT Leader Urges Caution in Law Schools' Use of GRE" and selecting the article from there. LSAC sure is scared.
Last edited by TheKingLives on Wed Aug 09, 2017 11:42 pm, edited 1 time in total.

User avatar
tuna_wasabi
Posts: 267
Joined: Tue Oct 27, 2015 5:16 pm

Re: GU and NU accept GRE - The Barbarians at the Gate for LSAC

Postby tuna_wasabi » Wed Aug 09, 2017 11:14 am

cavalier1138 wrote:
slurp wrote:ABA will be discussing more this fall. Apparently NU and GULC expressed that based on their own studies, the GRE is just as good, if not better, at predicting future law school success than the LSAT - which I wouldn't be surprised if proven true over the long run.


I would actually be pretty surprised by this, given how much easier it is to score well on the GRE. I wish that someone would actually show us the studies, because all we're hearing now is a general claim of "reliability" without being told about methodology, results, or sample size. If Northwestern just surveyed their students who took both in the past few years without controlling for anything else, then I'm much less inclined to trust that study.

I'm also deeply confused about how this opens doors in any meaningful way. Are students really deciding to not go to law school because they can't schedule around the appointed LSAT times?


I want to see these studies too. Didn't someone say that Harvard's study for this was flawed, due to unrepresentative sample? Did they use some sort of regression analysis to normalize the sample, or even better, to reflect the general test-taking population?

User avatar
Attorney-at-Birdlaw
Posts: 26
Joined: Thu Feb 25, 2016 6:56 pm

Re: GU and NU accept GRE - The Barbarians at the Gate for LSAC

Postby Attorney-at-Birdlaw » Wed Aug 09, 2017 11:38 am

cavalier1138 wrote:
slurp wrote:ABA will be discussing more this fall. Apparently NU and GULC expressed that based on their own studies, the GRE is just as good, if not better, at predicting future law school success than the LSAT - which I wouldn't be surprised if proven true over the long run.


I would actually be pretty surprised by this, given how much easier it is to score well on the GRE. I wish that someone would actually show us the studies, because all we're hearing now is a general claim of "reliability" without being told about methodology, results, or sample size. If Northwestern just surveyed their students who took both in the past few years without controlling for anything else, then I'm much less inclined to trust that study.

I'm also deeply confused about how this opens doors in any meaningful way. Are students really deciding to not go to law school because they can't schedule around the appointed LSAT times?


Joe Uhlman, What's So Great About the Lsat?, J. Kan. B.A., June 2017, at 36 (2017)

>The LSAT is a poor predictor of law school success. The test is designed to measure the skills the Law School Admissions Council believes are needed for future lawyers to succeed in law school, such as reading comprehension, analytical reasoning, and logical reasoning. However, at best, the LSAT only has a .56 correlation coefficient when predicting whether a student will fall into the top third of their class. More impartial metrics have found that the LSAT only predicts law school success with a 20% accuracy.

Recent experimental data score the GRE about the same as the LSAT in predicting law school success. There are no hard statistics about the GRE's prediction ability, because this academic year marks the first time in which students have been accepted to accredited law schools using the GRE. But the Educational Testing Service, the creators of the GRE, performed a study at the University of Arizona School of Law. The study found that GRE scores have a .55 correlation coefficient when predicting a student's law school success--or 1% worse than the LSAT.

User avatar
Attorney-at-Birdlaw
Posts: 26
Joined: Thu Feb 25, 2016 6:56 pm

Re: GU and NU accept GRE - The Barbarians at the Gate for LSAC

Postby Attorney-at-Birdlaw » Wed Aug 09, 2017 11:41 am

sparkytrainer wrote:


Paywall, can you post the text? But I can figure out why the lsat is scared- Gre is given every weekend, its instant scored, easier, and broader so you could apply to law school and phd programs with one test. If I was the lsat, I would be scared shitless.


Good, screw those people.

cavalier1138
Posts: 4295
Joined: Fri Mar 25, 2016 8:01 pm

Re: GU and NU accept GRE - The Barbarians at the Gate for LSAC

Postby cavalier1138 » Wed Aug 09, 2017 6:04 pm

Attorney-at-Birdlaw wrote:
cavalier1138 wrote:
slurp wrote:ABA will be discussing more this fall. Apparently NU and GULC expressed that based on their own studies, the GRE is just as good, if not better, at predicting future law school success than the LSAT - which I wouldn't be surprised if proven true over the long run.


I would actually be pretty surprised by this, given how much easier it is to score well on the GRE. I wish that someone would actually show us the studies, because all we're hearing now is a general claim of "reliability" without being told about methodology, results, or sample size. If Northwestern just surveyed their students who took both in the past few years without controlling for anything else, then I'm much less inclined to trust that study.

I'm also deeply confused about how this opens doors in any meaningful way. Are students really deciding to not go to law school because they can't schedule around the appointed LSAT times?


Joe Uhlman, What's So Great About the Lsat?, J. Kan. B.A., June 2017, at 36 (2017)

>The LSAT is a poor predictor of law school success. The test is designed to measure the skills the Law School Admissions Council believes are needed for future lawyers to succeed in law school, such as reading comprehension, analytical reasoning, and logical reasoning. However, at best, the LSAT only has a .56 correlation coefficient when predicting whether a student will fall into the top third of their class. More impartial metrics have found that the LSAT only predicts law school success with a 20% accuracy.

Recent experimental data score the GRE about the same as the LSAT in predicting law school success. There are no hard statistics about the GRE's prediction ability, because this academic year marks the first time in which students have been accepted to accredited law schools using the GRE. But the Educational Testing Service, the creators of the GRE, performed a study at the University of Arizona School of Law. The study found that GRE scores have a .55 correlation coefficient when predicting a student's law school success--or 1% worse than the LSAT.


Right. So I'd love to see what they mean when they say they "performed a study". We at least know LSAC's methodology; we know absolutely nothing about the Arizona study.

User avatar
Gordon_Cole
Posts: 34
Joined: Mon Aug 07, 2017 7:14 pm

Re: GU and NU accept GRE - The Barbarians at the Gate for LSAC

Postby Gordon_Cole » Thu Aug 10, 2017 1:13 am

Penn Law has stated they will not accept GRE in forseeable future http://www.thedp.com/article/2017/08/pe ... t-optional.

User avatar
Attorney-at-Birdlaw
Posts: 26
Joined: Thu Feb 25, 2016 6:56 pm

Re: GU and NU accept GRE - The Barbarians at the Gate for LSAC

Postby Attorney-at-Birdlaw » Thu Aug 10, 2017 3:52 pm

cavalier1138 wrote:
Right. So I'd love to see what they mean when they say they "performed a study". We at least know LSAC's methodology; we know absolutely nothing about the Arizona study.


I mean there are citations given in the footnotes, apologies for the text bomb:

http://online.wsj.com/public/resources/ ... rizona.pdf

Arizona Law provided LSAT scores and LGPA from its school files, and for the 12 student volunteers who had taken the GRE General Test prior to this study ETS supplied GRE scores from ETS records. For the remaining 66 student volunteers who had not already taken the GRE General Test, a special GRE test administration was arranged. Students were given a financial incentive of $50 to take the GRE General Test. In order to motivate students to use their best efforts on the GRE General Test, a student was informed that the student would receive an additional $50 payment if the student’s GRE percentile was not worse than twenty percentile points below the student’s LSAT percentile. All participants received this motivational incentive.

We employed standard statistical methods, including the calculation of summary
statistics (means and standard deviations). Because GRE-Q and GRE-V are adaptive multistage
tests (MSTs), we used an item response theory (IRT)-based method to calculate internal
consistency reliability for those subtests. For GRE-AW, we used classical test theory to
calculate internal consistency reliability. The GRE-AW reliability estimate is a lower bound on
the actual reliability of GRE-AW for reasons more fully described in the notes to Table 1. To
follow standard practice and to ensure comparability with LSAT, reliability was estimated based
on the testing population more fully described in the notes to Table 1. To measure validity, we
computed Pearson product moment correlation coefficients (i.e., r values), coefficients of
determination (i.e., R2 values) using ordinary least squares multiple regression, and contingency
tables. Because the correlation coefficients are based on data from students who were accepted
and then enrolled, they reflect validity for enrolled students. To ascertain the GRE validity for
the applicant pool for which the GRE General Test would be used to make admissions decisions,
we corrected the correlation coefficients for range restriction so that they would more accurately

The Validity of GRE® Scores for Predicting Academic Performance at the University of Arizona Law School reflect validity for the applicant pool. We calculated corrected correlation coefficients for the LSAT as well. These corrected correlation coefficients are estimates, because they assume that the variability in the scores of the applicants is known. Virtually everyone taking the GRE
General Test is interested in a graduate field other than law, so the correction that relies on the
variation of scores in the pool of GRE test takers is not fully clear. Nevertheless, adjusted
correlations are likely to be more accurate than unadjusted ones (see Schmidt & Hunter, 2015).

Another approach to adjusting for the admission of applicants based on their LSAT scores, but not their GRE scores, is to apply an explicit selection criterion to both tests. Specifically, we assumed that everyone enrolled in the law school could be thought of as an applicant pool to an even more selective institution. Under this assumption, we can set an explicit selection score by identifying the top three-quarters of this applicant pool based on LSAT scores, and separately, the top three-quarters based on GRE scores. Within each group of applicants, grade correlations can be computed and corrected for range restriction with the now known variance in the applicant population.

cavalier1138
Posts: 4295
Joined: Fri Mar 25, 2016 8:01 pm

Re: GU and NU accept GRE - The Barbarians at the Gate for LSAC

Postby cavalier1138 » Thu Aug 10, 2017 7:10 pm

Attorney-at-Birdlaw wrote:
cavalier1138 wrote:
Right. So I'd love to see what they mean when they say they "performed a study". We at least know LSAC's methodology; we know absolutely nothing about the Arizona study.


I mean there are citations given in the footnotes, apologies for the text bomb:


Ok, so that's a grand total of 78 students. And the LSAC studies on the correlation between 1L grades and LSAT/uGPA were conducted across the student bodies of 168 law schools.

Now, I'm in law school because I suck at math, but the Arizona study does not seem conclusive.

carsondalywashere
Posts: 439
Joined: Tue Apr 04, 2017 7:33 pm

Re: GU and NU accept GRE - The Barbarians at the Gate for LSAC

Postby carsondalywashere » Fri Aug 11, 2017 5:15 pm

Gordon_Cole wrote:Penn Law has stated they will not accept GRE in forseeable future http://www.thedp.com/article/2017/08/pe ... t-optional.

Lel, just for this cycle though.

Btw, love this thread title

User avatar
doggozeg
Posts: 32
Joined: Sun Jul 16, 2017 6:22 pm

Re: GU and NU accept GRE - The Barbarians at the Gate for LSAC

Postby doggozeg » Mon Aug 14, 2017 7:11 am

What's interesting is that the GRE is a much different test. There would be a whole different swath of people taking it I think. Also, MBA admissions take both GMAT and GRE, so those numbers could be worth using for predictions.




Return to “Law School Admissions Forum”

Who is online

Users browsing this forum: chrysippusofsoli, Contango, da489, lobsicle, steadfastness11, tada77, Will McAvoy and 6 guests