Do you also notice how LSAT sensitive USC is? Forum

[JD=Doctor]

I was looking at LSnumbers and if you got to the 167 mark you had a phenomenal shot at getting it
if you has a 166 your shot of getting in dropped by 85% unless your GPA was 3.9+

Talk about a school that loves trying to raise its medians...

[jamiLAW]

Yeah, it is a weird idea, but then again, I've never heard of a law school NOT concerned with its numbers.

I'm no expert, but it seems like those predictors don't tell the whole story; I was just talking to someone from Michigan who got a 158 on his LSAT. Admissions obviously puts a huge weight on numbers, but if your LORs/resume/PS didn't matter they wouldn't waste their time reading them.

[bdubs]

I was just talking to someone from Michigan who got a 158 on his LSAT.

That's how medians work. For every 2.95/173 kid, they need to consider a 3.97/159 kid.

That is generally not true, since the distribution of grades is not the same as the distribution of LSAT scores. The distribution of LSAT scores is close to normal, whereas GPAs tends to be closer to log normal (skewed toward the high end of the scale). Additionally, LSAT and GPA probably show some fairly significant degree of correlation.

Bottom line, there are lots of people who are just below the LSAT threshold with acceptable GPAs so your chances don't scale accordingly as you move down the LSAT ladder.

[r6_philly]

I was just talking to someone from Michigan who got a 158 on his LSAT.

That's how medians work. For every 2.95/173 kid, they need to consider a 3.97/159 kid.

That is generally not true, since the distribution of grades is not the same as the distribution of LSAT scores. The distribution of LSAT scores is close to normal, whereas GPAs tends to be closer to log normal (skewed toward the high end of the scale). Additionally, LSAT and GPA probably show some fairly significant degree of correlation.

Bottom line, there are lots of people who are just below the LSAT threshold with acceptable GPAs so your chances don't scale accordingly as you move down the LSAT ladder.

I can't speak for USC, but what does the distribution of the population has to do with the distribution of USC admit pool? The sample is artificially selected so it doesn't have to resemble the distribution of the population.

[bdubs]

I was just talking to someone from Michigan who got a 158 on his LSAT.

That's how medians work. For every 2.95/173 kid, they need to consider a 3.97/159 kid.

That is generally not true, since the distribution of grades is not the same as the distribution of LSAT scores. The distribution of LSAT scores is close to normal, whereas GPAs tends to be closer to log normal (skewed toward the high end of the scale). Additionally, LSAT and GPA probably show some fairly significant degree of correlation.

Bottom line, there are lots of people who are just below the LSAT threshold with acceptable GPAs so your chances don't scale accordingly as you move down the LSAT ladder.

I can't speak for USC, but what does the distribution of the population has to do with the distribution of USC admit pool? The sample is artificially selected so it doesn't have to resemble the distribution of the population.

Assuming their goals are in line with raising both medians, and their admits are drawn semi-randomly from the population they will not be considering kids with 3.97/159's because there are enough 3.97/164 kids waiting in the wings to play the USNWR ratings game.

Bottom line is I was differentiating between a median and an average. Nightrunner's description was more inline with an average, or the median of two normally distributed functions.

[turkeysub]

Do you also notice how LSAT sensitive USC every school is?

[Bumi]

Do you also notice how LSAT sensitive USC every school is?

http://berkeley.lawschoolnumbers.com/stats/0910/

Also OP, you ain't seen nothing until you've seen Emory.

http://emory.lawschoolnumbers.com/stats/0910/

[r6_philly]

Assuming their goals are in line with raising both medians, and their admits are drawn semi-randomly from the population they will not be considering kids with 3.97/159's because there are enough 3.97/164 kids waiting in the wings to play the USNWR ratings game.

Bottom line is I was differentiating between a median and an average. Nightrunner's description was more inline with an average, or the median of two normally distributed functions.

Don't they index the files base on a formula? I don't believe they are drawn "semi-randomly" to the files. They will have more applicants to choose from in one pool over the other pool for sure (GPA vs. LSAT) but I don't think their selection bias can be reflected by the distribution at all. I mean you can infer that but that would be inaccurate. The problem with this inference is you don't know 1. how many percent of each pool overlaps (By have both GPA/LSAT above median) and 2. how self-selection affects the sample population (the applicant pool, because applicants decide whether or not they apply based on a perceived bias instead of randomly applying). You just can't assume the group choosing to apply to USC resembles the overall population distribution. That is too much of a leap.

[r6_philly]

Do you also notice how LSAT sensitive USC every school is?

http://berkeley.lawschoolnumbers.com/stats/0910/

Also OP, you ain't seen nothing until you've seen Emory.

http://emory.lawschoolnumbers.com/stats/0910/

I love the Emory graph. They don't care about GPA.

[JOThompson]

Do you also notice how LSAT sensitive USC every school is?

Not so. The tiny spread between 25th/median/75th suggests an especially strong effort to climb the rankings. You don't see a range like 166/167/168 at every law school. Seems like the OP has a fair enough point, although you can certainly make the case about a number of other schools being just as or even more LSAT oriented.

[IAFG]

Do you also notice how LSAT sensitive USC every school is?

http://berkeley.lawschoolnumbers.com/stats/0910/

Also OP, you ain't seen nothing until you've seen Emory.

http://emory.lawschoolnumbers.com/stats/0910/

I love the Emory graph. They don't care about GPA.

my personal favorite: http://american.lawschoolnumbers.com/stats/0910/

162/163/164? enjoy american!

[r6_philly]

my personal favorite: http://american.lawschoolnumbers.com/stats/0910/

162/163/164? enjoy american!

Sandwich! So it their 25/median/75 162/163/164? That would be amazing.

[bdubs]

Don't they index the files base on a formula? I don't believe they are drawn "semi-randomly" to the files. They will have more applicants to choose from in one pool over the other pool for sure (GPA vs. LSAT) but I don't think their selection bias can be reflected by the distribution at all. I mean you can infer that but that would be inaccurate. The problem with this inference is you don't know 1. how many percent of each pool overlaps (By have both GPA/LSAT above median) and 2. how self-selection affects the sample population (the applicant pool, because applicants decide whether or not they apply based on a perceived bias instead of randomly applying). You just can't assume the group choosing to apply to USC resembles the overall population distribution. That is too much of a leap.

I was referring to the applicant pool distribution, not the admit pool distribution. Semi-randomly meaning that enough people want to go to USC to make its applicant pool resemble the general applicant pool for a given band of LSAT-GPA.

Obviously some people who are exceedingly overqualified will not apply, and those exceedingly under qualified will be discouraged from applying. That leaves you with all of the ranking crazed students who don't mind going to school in LA (probably a good number of people in that score band). I would assume that they strategically give out fee waivers to encourage the right students to apply as well.

If you assume there is no correlation between desire to go to school in CA and abnormal academic credentials, you are just left with the question of whether there are enough students randomly drawn from admittable pool to resemble the overall distribution of GPAs and LSATs. Considering that nearly 6600 people applied last year, I would say that it is safe to assume that.

[r6_philly]

I am pretty sure the percentage of applicants with 170+ AND 3.9+ applicants and 150- AND 2.8- applicants are not the same percentage as in the whole LSAC applicant pool. I am also pretty sure that the USC applicant pool has a mode somewhere near its published GPA/LSAT median, while the entirety of the LSAC applicant pool does not. Based on those 2 evidence, one can draw the conclusion (or at least not assume the opposite) that the USC pool does not resemble the entire LSAC applicant pool and should not be used as an statistical sample representing the entire pool.

That was my point, and you are welcome to disagree, but you will not convince me of otherwise.

[bdubs]

I am pretty sure the percentage of applicants with 170+ AND 3.9+ applicants and 150- AND 2.8- applicants are not the same percentage as in the whole LSAC applicant pool. I am also pretty sure that the USC applicant pool has a mode somewhere near its published GPA/LSAT median, while the entirety of the LSAC applicant pool does not. Based on those 2 evidence, one can draw the conclusion (or at least not assume the opposite) that the USC pool does not resemble the entire LSAC applicant pool and should not be used as an statistical sample representing the entire pool.

That was my point, and you are welcome to disagree, but you will not convince me of otherwise.

I think we just said the same thing.

[r6_philly]

Maybe now, but all the way back to last year we had a disagreement about something