Share Your Experiences, Read About Other Experiences. Please keep posts organized by school and expected year of graduation.

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

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

Unemployed
 Posts: 694
 Joined: Mon Feb 02, 2009 12:35 am
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
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 allstar in the NBA because you are tall. It's not that extreme of course but you get the idea.

T14_Scholly
 Posts: 419
 Joined: Mon Jan 18, 2010 8:46 pm
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 allstar 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.

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

T14_Scholly
 Posts: 419
 Joined: Mon Jan 18, 2010 8:46 pm
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: #gobroncos, allezallez21, andysimbi, april_ludgate, bas1, bbwannabe, BL2017, dylanca, gayfish, hayryem, Jack_Kelly, jazzy24, joooookchang, KENYADIGG1T, KylawRen, lqyzoe, lr1234, LSNlife, lumosfawkes, missmiabuttface, mwells56, Nik9639, rowdy, seraph9, Skadidas, Slawthschool, spectacles, ticket, tmcauliffe96, tony_dirko, usaorbust, wulfy10, WutSheOrder, yyzz, zzman and 111 guests