The most cogent anti-rankings article ever!

Share Your Experiences, Read About Other Experiences. Please keep posts organized by school and expected year of graduation.
User avatar
Panther7
Posts: 455
Joined: Mon Sep 28, 2009 5:34 pm

Re: The most cogent anti-rankings article ever!

Postby Panther7 » Thu Apr 08, 2010 8:12 pm

psychomohel wrote:
Panther7 wrote:As quoted by LSAC, 16% of you grades can be predicted by the LSAT, 4% by GPA. I don't think it's by coincidence that schools use the LSAT so heavily.


What does it mean to predict 16% of your grades?



the correlation factor, statistics thing... r squared, all that jazz. basically there is a 16% direct relationship between LSAT scores and LS GPA.

psychomohel
Posts: 181
Joined: Mon Nov 02, 2009 2:02 pm

Re: The most cogent anti-rankings article ever!

Postby psychomohel » Thu Apr 08, 2010 8:15 pm

Panther7 wrote:
psychomohel wrote:
Panther7 wrote:As quoted by LSAC, 16% of you grades can be predicted by the LSAT, 4% by GPA. I don't think it's by coincidence that schools use the LSAT so heavily.


What does it mean to predict 16% of your grades?



the correlation factor, statistics thing... r squared, all that jazz. basically there is a 16% direct relationship between LSAT scores and LS GPA.


I just looked it up, and apparently the median correlation of UGPA+LSAT and first year grades is .46, where 1 is a perfect correlation and 0 is no correlation at all.

User avatar
Panther7
Posts: 455
Joined: Mon Sep 28, 2009 5:34 pm

Re: The most cogent anti-rankings article ever!

Postby Panther7 » Thu Apr 08, 2010 8:17 pm

psychomohel wrote:
Panther7 wrote:
psychomohel wrote:
Panther7 wrote:As quoted by LSAC, 16% of you grades can be predicted by the LSAT, 4% by GPA. I don't think it's by coincidence that schools use the LSAT so heavily.


What does it mean to predict 16% of your grades?



the correlation factor, statistics thing... r squared, all that jazz. basically there is a 16% direct relationship between LSAT scores and LS GPA.


I just looked it up, and apparently the median correlation of UGPA+LSAT and first year grades is .46, where 1 is a perfect correlation and 0 is no correlation at all.


that's r factor, to get direct correlation or whatever you need to square that number, which gives you like .21, supposedly .16 of that is directly related to LSAT, the rest the UGPA.

User avatar
Unemployed
Posts: 699
Joined: Mon Feb 02, 2009 12:35 am

Re: The most cogent anti-rankings article ever!

Postby Unemployed » Thu Apr 08, 2010 8:23 pm

Panther7 wrote:
psychomohel wrote:I just looked it up, and apparently the median correlation of UGPA+LSAT and first year grades is .46, where 1 is a perfect correlation and 0 is no correlation at all.


that's r factor, to get direct correlation or whatever you need to square that number, which gives you like .21, supposedly .16 of that is directly related to LSAT, the rest the UGPA.


This.

Simply put, if the correlation coefficient is .46, then 21% of the variance in first year grades can be explained by the regression model (UGPA+LSAT).

bigben
Posts: 703
Joined: Sat Sep 29, 2007 2:44 pm

Re: The most cogent anti-rankings article ever!

Postby bigben » Thu Apr 08, 2010 8:39 pm

Unemployed wrote:
Panther7 wrote:
psychomohel wrote:I just looked it up, and apparently the median correlation of UGPA+LSAT and first year grades is .46, where 1 is a perfect correlation and 0 is no correlation at all.


that's r factor, to get direct correlation or whatever you need to square that number, which gives you like .21, supposedly .16 of that is directly related to LSAT, the rest the UGPA.


This.

Simply put, if the correlation coefficient is .46, then 21% of the variance in first year grades can be explained by the regression model (UGPA+LSAT).



The correlation is high enough to make the LSAT a remarkably good admissions tool. That you can find any single predictor with such a high correlation to something like law school grading is pretty amazing.

From the standpoint of trying to grant yourself some kind of competitive advantage, however, the correlation is dismally low. It makes almost no sense at all to factor it in.

As an illustration imagine just assuming that you will be an all-star in the NBA because you are tall. It's not that extreme of course but you get the idea.

User avatar
T14_Scholly
Posts: 416
Joined: Mon Jan 18, 2010 8:46 pm

Re: The most cogent anti-rankings article ever!

Postby T14_Scholly » Thu Apr 08, 2010 10:40 pm

bigben wrote:
Unemployed wrote:
Panther7 wrote:
psychomohel wrote:I just looked it up, and apparently the median correlation of UGPA+LSAT and first year grades is .46, where 1 is a perfect correlation and 0 is no correlation at all.


that's r factor, to get direct correlation or whatever you need to square that number, which gives you like .21, supposedly .16 of that is directly related to LSAT, the rest the UGPA.


This.

Simply put, if the correlation coefficient is .46, then 21% of the variance in first year grades can be explained by the regression model (UGPA+LSAT).



The correlation is high enough to make the LSAT a remarkably good admissions tool. That you can find any single predictor with such a high correlation to something like law school grading is pretty amazing.

From the standpoint of trying to grant yourself some kind of competitive advantage, however, the correlation is dismally low. It makes almost no sense at all to factor it in.

As an illustration imagine just assuming that you will be an all-star in the NBA because you are tall. It's not that extreme of course but you get the idea.


Looks like your username isn't gonna be so cool when big Ben is in jail.

User avatar
Panther7
Posts: 455
Joined: Mon Sep 28, 2009 5:34 pm

Re: The most cogent anti-rankings article ever!

Postby Panther7 » Fri Apr 09, 2010 7:03 am

T14_Scholly wrote:Looks like your username isn't gonna be so cool when big Ben is in jail.


--ImageRemoved--

User avatar
T14_Scholly
Posts: 416
Joined: Mon Jan 18, 2010 8:46 pm

Re: The most cogent anti-rankings article ever!

Postby T14_Scholly » Fri Apr 09, 2010 12:36 pm

No one cares about that Big Ben.




Return to “Law School Acceptances, Denials, and Waitlists”

Who is online

Users browsing this forum: 20171lhopeful, Google Adsense [Bot], tinafeyclone and 5 guests