GPA weight corresponding to school? Forum

[cavalier1138]

I never claimed that this is true for every law school. This is just for our school’s own law school, which knows our undergrads very well. Our undergrads career service is one of the best in the country, and the one I talked to is the director of our law career program, who used to work for yls and works directly with our law school, so it is unlikely my data is off. 0.2 is more of an estimate, so let’s just just say that our law school’s median gpa is roughly higher than the 95%tile of the undergrads gpa and our law school admissions synpathesize with us and believe that the top 25% or so should already suffice, though that’s about 0.2 lower than their median and evenlower than their 25%tile

Do you know what percentage of the law school's students also went to the undergrad?

[zwcai]

I would have expected most of the T-14 law schools to have a median GPA above the 90th percentile of their undergraduate school. The mean ug gpa at uva is 3.42 for females, 3.36 for men. The law school's 25th, 50th and 75th percentiles are 3.59, 3.89, 3.97.

The idea that a law school would accept the entire top 25th percent of graduates seems really unusual. The number of applicants would be hugs (assuming this was a T-14 school) because many of the class wouldn't be competitive elsewhere. If they were accepted, the US News algorithms would then punish the law school.

I’m sure they still have LSAT requirements or maybe they accept juniors. It sounds more like a state school to me or maybe BU,Notre Dame, etc. Every Ivy school except Princeton is known more for grade inflation I think. Not sure about Cornell because I always forget it.

But zwcai's post strongly suggested that the law school takes the entire top quarter of their university's undergrads - not only those members of the top quarter who kill it on the LSAT (in which case they could simply be admitted as regular splitters - they wouldn't even need a GPA boost).

I agree that the law school is unlikely to be in the T13, or even the T20.

This is NOT what I’m saying here. Top 25% of the class is way more than who actually applies to law schools each year from my school. Again, by your definition, only people who are right at the 75th%tile from my undergrad apply and get into our law school, which is definitely not the case. What I’m saying here is that having a 75th%tile gpa in our OWN undergrad would not put you at a great disadvantage at our OWN law school given that you have median LSAT, even though it is below the law school median by about 0.2. Of course, lots of people with great gpas from my undergrad apply and get into our law school, so it’s not like only the bad kids get in. I’m only saying that our law school looks at its own undergrads more favorably and does not expect as high of a gpa from them. And no, it’s not a state school. It’s indeed a t13, a pretty good one too.

[kellyjohnson]

The idea that a law school would accept the entire top 25th percent of graduates seems really unusual. The number of applicants would be hugs (assuming this was a T-14 school) because many of the class wouldn't be competitive elsewhere. If they were accepted, the US News algorithms would then punish the law school.

First, not all of the entire top 25% of graduates would apply to the law school.Only a fraction of the top 25% of the school are even interested in pursuing law. Second, not all of the top 25% will be above median LSAT, and so not all would be accepted. All he's saying is there is flexibility on GPA, not on LSAT.

Second, not everyone in that top 25 percent has below median GPA. It could be that the top 12% has above median GPA, and then 12-25% has a below median GPA. In which case, accepting everyone in the top 25% would barely affect the median, since more or less half of the acceptances would be above and half would be below. Now, possibly there are more accepted applicants in the 12-25% range than in the 1-12% range (but it could also be opposite given that its more likely those in the 1-12% range would have LSAT scores above median).

Third, half of a law school's students have below median GPA. So there are lots of spaces for below median GPA students. All this policy would do is possibly displace other students (from other UG) that have below median GPA.

It is quite possible to have this policy without really affecting medians.

[zwcai]

The idea that a law school would accept the entire top 25th percent of graduates seems really unusual. The number of applicants would be hugs (assuming this was a T-14 school) because many of the class wouldn't be competitive elsewhere. If they were accepted, the US News algorithms would then punish the law school.

First, not all of the entire top 25% of graduates would apply to the law school.Only a fraction of the top 25% of the school are even interested in pursuing law. Second, not all of the top 25% will be above median LSAT, and so not all would be accepted. All he's saying is there is flexibility on GPA, not on LSAT.

Second, not everyone in that top 25 percent has below median GPA. It could be that the top 12% has above median GPA, and then 12-25% has a below median GPA. In which case, accepting everyone in the top 25% would barely affect the median, since more or less half of the acceptances would be above and half would be below. Now, possibly there are more accepted applicants in the 12-25% range than in the 1-12% range (but it could also be opposite given that its more likely those in the 1-12% range would have LSAT scores above median).

Third, half of a law school's students have below median GPA. So there are lots of spaces for below median GPA students. All this policy would do is possibly displace other students (from other UG) that have below median GPA.

It is quite possible to have this policy without really affecting medians.

^This. You are ten times better than me at explaining stuff. This is exactly what I was thinking but very poorly conveyed.

[albanach]

First, not all of the entire top 25% of graduates would apply to the law school.Only a fraction of the top 25% of the school are even interested in pursuing law. Second, not all of the top 25% will be above median LSAT, and so not all would be accepted. All he's saying is there is flexibility on GPA, not on LSAT.

Second, not everyone in that top 25 percent has below median GPA. It could be that the top 12% has above median GPA, and then 12-25% has a below median GPA. In which case, accepting everyone in the top 25% would barely affect the median, since more or less half of the acceptances would be above and half would be below. Now, possibly there are more accepted applicants in the 12-25% range than in the 1-12% range (but it could also be opposite given that its more likely those in the 1-12% range would have LSAT scores above median).

Third, half of a law school's students have below median GPA. So there are lots of spaces for below median GPA students. All this policy would do is possibly displace other students (from other UG) that have below median GPA.

It is quite possible to have this policy without really affecting medians.

I'm not sure you're getting my point. Only a fraction of a school being interested in law school would still be a huge number of applicants for any top law school. Let's consider the T-100 to be law schools that might have reasonable outcomes. There are over 1,500 four year colleges in the United States.

Sure, not everyone in the top 25% is going to be below median. The poster said it was only 20% of them (median is 95th percentile).

Let's look at UVa again as an example. Here's what a cycle looks like for a candidate with median lsat and gpa:



Now, if we drop their gpa by 0.2 (the boost that was mentioned) their cycle looks like this:



They've gone from a decent chance at admission everywhere in the lower T14 to a much less positive cycle. So graduates that would otherwise naturally have spread out among the peer schools are now all going to your institution because their boost would make them technically above both medians.

I'm pretty confident you'd see something similar at any decently ranked non-regional school.

If you fill up the bottom of the GPA percentiles with your class, you now can't offer as many places to splitters to boost your GPA median. All this with relatively small class sizes of 1-300 means that any significant boost given to ugpa for students from the same school would almost certainly start to impact the school's medians really quickly.

[kellyjohnson]

Yeah, but not every student in the top 25% is going to be at a 169 / median lsat, which your argument assumes. Some will themselves be splitters (with higher LSAT), and some will have lower LSAT (so the boost very well may not help them, as they'd get rejected anyway). You are assuming a lot in your argument. We dont know the profile or numbers of those that typically apply.

Also, for all we know this practice *does impact* median GPA's. It could be the schools GPA is, say, 0.05 - 0.1 GPA points lower as it stands right now than it would be if this were not the policy.

[zwcai]

I really wish to remain anonymous so I will point out some errors in your estimation without leaking information about my school as much as possible. A fraction of the class is a very small number. A fraction of the top 25% is even smaller, because most of these kids go into finance. This only applies to our own law school, so comparing this with possibilities at other law schools is kinda pointless. Finally, with that being said, a median LSAT at our law school combined with a not so good gpa by our law school's standard still gives you a pretty good shot at almost all the lower t14's.

[albanach]

Yeah, but not every student in the top 25% is going to be at a 169 / median lsat, which your argument assumes.

My argument is that when you increase by four hundred percent the number of competitive applicants (top 5 percent of graduating class to top 25%) you end up with a very large number of applicants who are vastly more competitive at your school than at peer schools.

The only assumption here is that the number of students with competitive LSAT scores is fairly evenly distributed among the top 25% of the class. I make that assumption because we know LSAT has a higher correlation to law school performance than undergraduate GPA, therefore we know that, at least to an extent, LSAT and uGPA don't correlate that tightly (or there'd be no need for the LSAT).


I certainly don't assume, as you suggest, that everyone in the top 25% has a competitive LSAT.

[QContinuum]

You are assuming a lot in your argument. We dont know the profile or numbers of those that typically apply.

The core of zwcai's claim is that undergrads from their college can get in to their law school with a below-median GPA, without compensating with a high LSAT. zwcai claims that this happens, because they and their classmates have their GPAs boosted by 0.2, so that a below-median GPA - say, a 3.7 - would be treated as if it were an above-median GPA - say, a 3.9.

It's not disputed by anyone that undergrads from zwcai's college can get in to their law school with a below-median GPA and an above-median LSAT. That's the definition of a splitter. When a splitter gets in, that doesn't mean they're getting a GPA "boost" - it just means they're using their high LSAT score to compensate for their lower GPA.

A fraction of the class is a very small number. A fraction of the top 25% is even smaller, because most of these kids go into finance.

Sure, a fraction of the top 25% may be small relative to the size of the entire undergraduate class, but it's still going to be pretty significant relative to the size of the law school class. UMich, for instance, has an undergraduate enrollment of 30,000. Its J.D. class has a total enrollment of 931. Its law school is 1/30th - that's 3.1% - the size of its college.

[zwcai]

You are assuming a lot in your argument. We dont know the profile or numbers of those that typically apply.

The core of zwcai's claim is that undergrads from their college can get in to their law school with a below-median GPA, without compensating with a high LSAT. zwcai claims that this happens, because they and their classmates have their GPAs boosted by 0.2, so that a below-median GPA - say, a 3.7 - would be treated as if it were an above-median GPA - say, a 3.9.

It's not disputed by anyone that undergrads from zwcai's college can get in to their law school with a below-median GPA and an above-median LSAT. That's the definition of a splitter. When a splitter gets in, that doesn't mean they're getting a GPA "boost" - it just means they're using their high LSAT score to compensate for their lower GPA.

A fraction of the class is a very small number. A fraction of the top 25% is even smaller, because most of these kids go into finance.

Sure, a fraction of the top 25% may be small relative to the size of the entire undergraduate class, but it's still going to be pretty significant relative to the size of the law school class. UMich, for instance, has an undergraduate enrollment of 30,000. Its J.D. class has a total enrollment of 931. Its law school is 1/30th - that's 3.1% - the size of its college.

I’m just going to say that my college is tiny. The number of applicants from my undergrad is about 1% of the total number of applicants our law school gets. No all of them are going to qualify for sure, and not all of those who qualify will matriculate, so I’m going to guess they represent maybe 5% of the law school population? But you get the gist of it.

[Npret]

You are assuming a lot in your argument. We dont know the profile or numbers of those that typically apply.

The core of zwcai's claim is that undergrads from their college can get in to their law school with a below-median GPA, without compensating with a high LSAT. zwcai claims that this happens, because they and their classmates have their GPAs boosted by 0.2, so that a below-median GPA - say, a 3.7 - would be treated as if it were an above-median GPA - say, a 3.9.

It's not disputed by anyone that undergrads from zwcai's college can get in to their law school with a below-median GPA and an above-median LSAT. That's the definition of a splitter. When a splitter gets in, that doesn't mean they're getting a GPA "boost" - it just means they're using their high LSAT score to compensate for their lower GPA.

A fraction of the class is a very small number. A fraction of the top 25% is even smaller, because most of these kids go into finance.

Sure, a fraction of the top 25% may be small relative to the size of the entire undergraduate class, but it's still going to be pretty significant relative to the size of the law school class. UMich, for instance, has an undergraduate enrollment of 30,000. Its J.D. class has a total enrollment of 931. Its law school is 1/30th - that's 3.1% - the size of its college.

I’m just going to say that my college is tiny. The number of applicants from my undergrad is about 1% of the total number of applicants our law school gets. No all of them are going to qualify for sure, and not all of those who qualify will matriculate, so I’m going to guess they represent maybe 5% of the law school population? But you get the gist of it.

I’m sure you have accurate info about your own school. I was only thinking of Princeton because I guess their undergrads get or used to get slight upgrades in GPA with T14 admissions. I’m sure other schools can do the same for their own undergrads whether or not it’s well known outside the school. I think it can be done without dropping medians. Im not going to speculate further as to your school.

[kellyjohnson]

You are assuming a lot in your argument. We dont know the profile or numbers of those that typically apply.

The core of zwcai's claim is that undergrads from their college can get in to their law school with a below-median GPA, without compensating with a high LSAT. zwcai claims that this happens, because they and their classmates have their GPAs boosted by 0.2, so that a below-median GPA - say, a 3.7 - would be treated as if it were an above-median GPA - say, a 3.9.
.

Yes, but my point is that 90% of the top 25% may have an lsat >169 and be splitters, in which case his entire analysis falls apart. Same if 90% are below 169. we dont know enough to make the conclusions hes making.

The only assumption here is that the number of students with competitive LSAT scores is fairly evenly distributed among the top 25% of the class. I make that assumption because we know LSAT has a higher correlation to law school performance than undergraduate GPA, therefore we know that, at least to an extent, LSAT and uGPA don't correlate that tightly (or there'd be no need for the LSAT).

And there it is, the hand-waving on which your whole analysis relies. You have no idea how correlated LSAT and uGPA are for this particular school, nor for this particular sub-population within this particular school. (For those studying for the LSAT, the flaw here is: "Because something is true for a population, we can assume that thing is also true for each subgroup within that population.")

To the OP: Trust the person at your school, not the people on this forum.

[albanach]

To the OP: Trust the person at your school, not the people on this forum.

You recognize you are suggesting someone trust the person in undergraduate careers counselling, an area legend for its inability to comprehend law school admissions? As I said previously, if this came from the law school admissions, that would be one thing, but that is not the case.

[QContinuum]

Yes, but my point is that 90% of the top 25% may have an lsat >169 and be splitters, in which case his entire analysis falls apart. Same if 90% are below 169. we dont know enough to make the conclusions hes making.

It's extremely unlikely that 90% of the top 25% of undergrads at zwcai's school would have a 97.4+% LSAT (the percentile corresponding to a 170). It's so unlikely I'm comfortable suggesting it's practically impossible.

The only assumption here is that the number of students with competitive LSAT scores is fairly evenly distributed among the top 25% of the class. I make that assumption because we know LSAT has a higher correlation to law school performance than undergraduate GPA, therefore we know that, at least to an extent, LSAT and uGPA don't correlate that tightly (or there'd be no need for the LSAT).

And there it is, the hand-waving on which your whole analysis relies. You have no idea how correlated LSAT and uGPA are for this particular school, nor for this particular sub-population within this particular school. (For those studying for the LSAT, the flaw here is: "Because something is true for a population, we can assume that thing is also true for each subgroup within that population.")

You and zwcai err in placing great weight on the fact that only a "sub-population" of undergrads are interested in law. But there's nothing special about that "sub-population": Everyone taking the LSAT nationwide is in that same "sub-population" of law school aspirants. It's not like there are premeds and b-school aspirants taking the LSAT at other schools.

LSAT and uGPA don't correlate tightly, nor is there any reason to remotely suspect a significantly different correlation at zwcai's school.

[albanach]

I’m just going to say that my college is tiny. The number of applicants from my undergrad is about 1% of the total number of applicants our law school gets. No all of them are going to qualify for sure, and not all of those who qualify will matriculate, so I’m going to guess they represent maybe 5% of the law school population? But you get the gist of it.

It's worth pointing out that none of the T-14 have a tiny college.

Yale - 5,453
Stanford - 7,062
Harvard - 6,700
Chi - 6,264
Colubia - 8,410
NYU - 28,799
Penn - 10,496
Mich - 29,821
UC Berk - 30,853
UVA - 16,655
Duke - 6,994
NW - 8,353
Cortnell - 15,182
GULC - 7,636

Yale has the smallest number of undergraduates; however GPAs there are reported to be inflated, with 30% of the class commanding a 3.7 or above.

Stanford has a larger school, and a bit less grade inflation, but nothing out of the ordinary. Their median in 2011 was 3.57 and was on an upward trajectory.

Harvard has a small class, but their median grade is A-

Chicago isn't small, but is commonly reported to have little in the way of grade inflation. However, the 2006 mean award of 3.35 was within 0.1 of places like Harvard. I find it interesting that the school doesn't seem to publish any stats which, if there had been no grade inflation in the intervening decade, would doubtless be helpful to students communicating with employers. Either way, it doesn't seem to be out of whack with other T-14 schools like UVA (2018 male average 3.36, female average 3.42). Indeed the 2006 figure puts Chicago's mean GPA above UVA's.

Maybe one of the other schools has a dramatically lower uGPA? I don't have time right now to search for them all.

[QContinuum]

Maybe one of the other schools has a dramatically lower uGPA? I don't have time right now to search for them all.

Thanks albanach for the yeoman's work. You've conclusively ruled out Yale, Stanford, Harvard, and Chicago. zwcai says their school isn't a "state school," so that rules out Penn, Mich, Berkeley, and Virginia. The school is described as small, so that rules out NYU and Cornell (NYU is massive, approaching 30k undergrads, and Cornell, at >15k undergrads, can't remotely be characterized as small). zwcai also says that the school is "a t13, a pretty good one," so that rules out Georgetown.

We're left with Columbia, Duke, and Northwestern. A report on the Columbia Spectator indicates that in the early 2010s, it was common for at least some Columbia departments to award undergrads >60% As or A-s (up to 75% in Music Humanities). This doesn't sound like grade deflation. Duke reportedly has seen "some of the sharpest grade inflation over the past half century," according to a 2016 report. The average GPA at Northwestern in 2015 was allegedly 3.48, higher than Chicago.

[Wubbles]

Maybe one of the other schools has a dramatically lower uGPA? I don't have time right now to search for them all.

Thanks albanach for the yeoman's work. You've conclusively ruled out Yale, Stanford, Harvard, and Chicago. zwcai says their school isn't a "state school," so that rules out Penn, Mich, Berkeley, and Virginia. The school is described as small, so that rules out NYU and Cornell (NYU is massive, approaching 30k undergrads, and Cornell, at >15k undergrads, can't remotely be characterized as small). zwcai also says that the school is "a t13, a pretty good one," so that rules out Georgetown.

We're left with Columbia, Duke, and Northwestern. A report on the Columbia Spectator indicates that in the early 2010s, it was common for at least some Columbia departments to award undergrads >60% As or A-s (up to 75% in Music Humanities). This doesn't sound like grade deflation. Duke reportedly has seen "some of the sharpest grade inflation over the past half century," according to a 2016 report. The average GPA at Northwestern in 2015 was allegedly 3.48, higher than Chicago.

Had to correct you on Penn being a state school

Anyways, I actually would guess Chicago has the least inflation of these schools and is where OP is at. Or OP is in a specific program at another small t14 that deflates, which is very much possible, but basically no one would be aware of it.

[QContinuum]

Had to correct you on Penn being a state school

Anyways, I actually would guess Chicago has the least inflation of these schools and is where OP is at. Or OP is in a specific program at another small t14 that deflates, which is very much possible, but basically no one would be aware of it.

Noted re Penn! My bad. Still, it's probably not Penn, as Penn has >10k undergrads so isn't small by any stretch.

You're probably right that it's Chicago. I wouldn't necessarily say Chicago deflates, but it at least doesn't inflate as much as some of the other T13s' affiliated colleges do.

[albanach]

You're probably right that it's Chicago. I wouldn't necessarily say Chicago deflates, but it at least doesn't inflate as much as some of the other T13s' affiliated colleges do.

If it is Chicago (and based on the description and analysis, that seems most likely) is be astonished if there was a boost of 0.2. Perhaps some individual majors might see a boost like that, but applied to the whole school? This sounds like one of those things the administration says to stop undergrads who are paying $55k/year in tuition, revolting in anger when they learn they are at a disadvantage to peer schools.

Most folk here have long since accepted that they cannot compete against a 4.0 from a community college, regardless of the difficulty of their course or prestige of their school.

[QContinuum]

You're probably right that it's Chicago. I wouldn't necessarily say Chicago deflates, but it at least doesn't inflate as much as some of the other T13s' affiliated colleges do.

If it is Chicago (and based on the description and analysis, that seems most likely) is be astonished if there was a boost of 0.2. Perhaps some individual majors might see a boost like that, but applied to the whole school? This sounds like one of those things the administration says to stop undergrads who are paying $55k/year in tuition, revolting in anger when they learn they are at a disadvantage to peer schools.

Most folk here have long since accepted that they cannot compete against a 4.0 from a community college, regardless of the difficulty of their course or prestige of their school.

Agreed. Granted, it's well-known that Yale, for instance, does favor Yale College applicants. So I wouldn't be surprised if there was some kind of "Chicago Preference" in effect. But I seriously doubt, as you say, that any boost would be implemented as an across-the-board, school-wide 0.2 adjustment.

Overall, of course, the schools that don't do grade inflation are doing their undergrads a grave disservice. People may "know" about Princeton's grade deflation, but if I was choosing which college to attend, I'd take Yale over Princeton any day of the week and twice on Sundays. GPA matters a lot and I'm not convinced Princetonians get anywhere near enough of a GPA adjustment to make up for their grade deflation - people will likely just be like, "oh, a 3.5 at Princeton's better than a 3.5 anywhere else" - without recognizing the Princeton 3.5 as a 3.7. And at least Princeton's pretty well-known for its grade deflation. Prior to this whole thread I had no clue Chicago did anything different grades-wise than its peer colleges.

[AJordan]

Short answer: no. Because of the reliance on reportable numbers a GPA is a GPA.

Long answer: it depends on both institutions. HLS may have more respect for a Harvard degree but it’s not like they’re going to prioritize a 3.5 from a tough liberal arts school over a 3.95 from random state.

All things equal it’s another factor to consider. It could be double-sides I guess. They might expect your written materials to be better and, as such, more likely to be disappointed in their content. This is all speculation, though. Everyone with the numbers gets considered minus some glaring flaw. It’s probably not worth getting hung up about unless there’s some sort of direct connection between the undergrad and the law school.

I'd like to cite the rest of this thread as evidence for my initial response.

[albanach]

I'd like to cite the rest of this thread as evidence for my initial response.

Indeed. That said, while it can be frustrating when there's a challenge to the perceived wisdom. it's helpful every now and again to think about this and take a look at the facts. I'm more confident now than I was before the thread started, even though I would have given the same answer as you. I truly hadn't considered if any of the T-14 had a tiny undergraduate college (turns out they don't but it could have been influential if they had). And, while there's obviously disparity in grades between the T-14 colleges give their students, I'm pretty comfortable that none are handing out GPAs on a B median or anything totally out of whack like that.

If I managed a site like lawschoolnumbers, I think I'd consider enforcing users to select their UG institution so that they can track any statistically significant GPA boost. I think that would be useful information to make available.

[AJordan]

I'd like to cite the rest of this thread as evidence for my initial response.

Indeed. That said, while it can be frustrating when there's a challenge to the perceived wisdom. it's helpful every now and again to think about this and take a look at the facts. I'm more confident now than I was before the thread started, even though I would have given the same answer as you. I truly hadn't considered if any of the T-14 had a tiny undergraduate college (turns out they don't but it could have been influential if they had). And, while there's obviously disparity in grades between the T-14 colleges give their students, I'm pretty comfortable that none are handing out GPAs on a B median or anything totally out of whack like that.

If I managed a site like lawschoolnumbers, I think I'd consider enforcing users to select their UG institution so that they can track any statistically significant GPA boost. I think that would be useful information to make available.

I certainly agree with your first point. As to the second, I think we've been spoiled by LSN as a resource. It's probably impossible to monetize it and keep it informative enough to be useful. Then we had attack of the spambots and its utility reduced drastically. RIP LSN.

[Pompey]

I agree that the law school is unlikely to be in the T13, or even the T20.

I have a feeling it's Berkeley, actually.

I would say U Chicago.