Low GPA from Harvard?

mcat2 · May 31, 2009, 8:28pm

BB: As far as I know, every course can be taken P/NP at Brown, if the student chooses to do so. This would encourage students to take only those courses non-P/NP which they believe they will likely get a good grade. It may be due to this reason that Brown students have higher GPAs than almost all other Ivies. For example, according to what BDM posted, the grade inflation/deflation adjusting factors for UPenn, Princeton, Harvard, etc., have a much more negative adjusting factor than that for Brown.

I would imagine that if the students at my child’s school are allowed to do this, their GPAs will likely be much higher. At my child’s school, I heard some extreme left-brainers or right brainers are forced to struggle in some of the required courses, and possibly taking a hit for the GPA, due to the distribution requirement; all of these courses have to be taken with a letter grade (and you can not AP any of these out.)

Although my child is currently not on the prelaw track (he is on the premed track as of today, but who knows whether he will change his mind in the end), he said that in the past semester, he spent more time on his non-premed classes than his premed ones last semester. I could imagine that he could allocate much more study time for his premed pre-req’s if he took one or more of his non-premed ones P/NP.

White_Rabbit · May 31, 2009, 8:53pm

You don’t even have to see your advisor. Go to the OCS and buy the grids for $3 or something like that. They are very helpful.

mcat2 · May 31, 2009, 9:01pm

What is OCS? Is it only for Harvard’s student?

White_Rabbit · May 31, 2009, 11:01pm

Office of Career Services, yes

Lergnom · May 31, 2009, 11:58pm

I thought it was Officer Candidate School. Fort Benning. And here I thought Harvard didn’t respect the military.

bluebayou · June 1, 2009, 12:12am

Sorry, but I’m still struggling to see the logic. Why not use the SAT (or tax return) since it will correlate just as well (H has the highest SAT scores in the land, as LSAT.)

mcat: I believe Brown also has a liberal drop policy, as in drop any class late in the term w/o penalty or record on transcript. But, regardless, I can’t fathom how Brown is DEflated according to bdm…

bluedevilmike · June 1, 2009, 1:02am

We could use the SAT or tax return, but we don’t have that information for this particular population. More importantly, the LSAT is the standardized test used to measure this particular population, so it’s the logical one to use.

I suspect we’d get similar results with SATs, although I don’t know about socioeconomic demographics. But I don’t know the SAT scores for pre-law students specifically.

Brown, at 0.16 standard deviations harder than the mean, is the most inflated Ivy. It’s easier than UCSD (0.21), Berkeley (0.65), etc. But the overall point is that Ivies generally (Penn, 1.49; Harvard, 0.93; Princeton 0.81) are harder than publics such as Berkeley.

sakky · June 1, 2009, 3:03am

I still don’t buy it, with my greatest concern being the tail-end risk. Let’s be perfectly honest - it’s practically impossible to actually flunk out of Harvard or Princeton. Even their worst students will still pass. Not so at publics like Berkeley, where flunking out is a distinct possibility. Those who do will be excluded from the law school applicant pool for the simple reason that they can’t even complete their bachelor’s degrees.

I think the best that can be said is that while it may perhaps be harder to get top grades at Harvard or Princeton as opposed to a public - although I’m still not convinced that is the case - it is also clearly infinitely easier to avoid truly bad grades at the former schools for the simple reason that such grades are never handed out in the first place.

bluebayou · June 1, 2009, 12:28pm

bdm:

If I understand correctly, what you are suggesting is that since H has cornered the market on the “smartest” kids – as defined by one ~4 hour test – by definition, it should have the highest average grades. And, any other college which has a lower mean LSAT score (i…e, all of 'em), should in fact have lower gpa’s, or it is considered inflated, correct? Thus, H could give out B’s like candy but as long as they make earning A’s difficult, it’s mean gpa will remain below Brown’s, and it will be grade DEflated relative to Brown?

No offense intended towards H or B folks, just trying to understand the gpa-lsat equating concept.

mcat2 · June 1, 2009, 1:04pm

BB: I believe this is the essence of what BDM was saying.

I believe the real issue is: For the professional school admission purpose, what should be considered as a “fair” system?

Even for college admission, a student with a very good GPA is sometimes, fairly or unfairly, labeled with negative terms like “grind”, “bookworm”, etc. There is some truth that the “quality” of the student may be measured by the standardized test to some extent, but it is also true that we can not totally discard the value of GPA as a reasonable merit measuring method. In addition, the quality of the EC can also tell us a lot about the student so it is also an important factor. (BTW, regarding the EC, I sometimes feel that because the FAMILY of the student is more resourceful, the student can have more ready access to a good EC. This is especially true during the undergraduate admission cycle, when many students are still too young/inexperienced to be independent of their parents.)

What we can not agree upon is the weight of each of these factors (and I think we can never agree on this unfortunately).

I also believe that a student who is good (by nature or by nurture/grind?) at standardized tests always has a great advantage on many fronts, esp. educational opportunities.

Lergnom · June 1, 2009, 1:10pm

I also disagree. First, you need to accept the LSAT as a metric independent of its use in law school admissions, which is dubious because most kids don’t apply to law school and kids who take the LSAT may do poorly in, say, math. Second, you need to define the correlation between the LSAT, the individual and grades - plus add in the stuff I mentioned above, such as the relative inflexibility of points in GPA versus LSAT and how that skews the apparent distribution.

If one means the Harvard student now going through life as Berkeley student would do better at Berkeley, that’s not implied by the data; it merely says that if you use the LSAT to set a metric, then you can rank applicants from schools by their expected grades. That has no causative meaning in any other context. If you’re saying that a kid from Harvard, a law school applicant, who somehow magically took the same curriculum at Berkeley would do better that implies the Berkeley student would do worse at Harvard. The data implies the opposite, that Harvard inflates all grades.

As I read the metric, it allows law schools to look at individuals from a school in context of their school. If a kid from Harvard does below Harvard average, that could be a worse performance than a kid from elsewhere who outperforms his or her own school’s expected metric. It means that kids don’t have to worry about grade inflation / deflation for their school but about how they perform relative to their school’s metric. One kid can thus be a much higher over-performer with the same record as a kid from another school because that other kid may actually be worse than expectations from that school’s metric.

jonri · June 1, 2009, 5:53pm

The “metric” ?–is that really want you meant?-- is used for two purposes: the first is to give the LSs some sense of the grading system at X college. Kids DO have to worry about grade inflation at their schools! It is especially useful in giving law schools some idea of the grading system at colleges with which they are unfamiliar. While only one student from Old Widget Liberal Arts College may be applying to a given LS, the data compiled by LSDAS allows the LS to find out the median LSAT of Widgets applying to all law schools and how the applicant’s gpa stacks up against all of the applicants from Old Widget to ALL LSs. So, the law school for the U of Wyoming may never have received an application from a Widget before, but when it does, it will know a lot more info about what a 3.4 at Old Widget “means” than it would if the LSDAS didn’t exist.

The second purpose is to tell LSs how the applicant stacks up against other applicants from that college.

The LSDAS (Law School Data Assembly Service) uses LSAT and gpa’s of applicants to LS in large part because they are available. It knows the actual LSAT scores of each applicant. It receives the actual transcripts of those applicants and it attempts to standardize the grading systems as much as it can. You can, in theory, come up with other measures that might work better, but there’s no way to insure that the LSDAS would get actual, unadulterated data. That’s why LSDAS uses them bluebayoo–there’s no other reason.

LOTS of colleges are SAT optional schools. Others accept the ACT in lieu of the SAT. If LSDAS used SAT scores, how would it account for schools at which SATs are optional? Schools like Bowdoin and Mt. Holyoke–both fine schools–don’t require SAT scores and thus there is no way that LSDAS could acquire SAT scores from all LS applicants. Another way colleges “massage” SAT scores is to give summer or spring admissions to lower-scoring students. That way they aren’t included in USNews’ data–even though those students are enrolled in the college. Assuming that these students get lower grades in the aggregate than the students admitted in the usual fashion, comparing the reported SATs with the actual gpa’s would make the school look tougher than it is. And, of course, some kids took the ACTs, not the SATs.

As for gpa’s–many colleges don’t release that information. Even among those that do, the methodology varies a lot. LSDAS can’t standardize gpa’s perfectly, but it certainly does so more effectively than colleges do. For example, some colleges only count the new higher grader in gpa for repeated courses. Others count both grades. If it can figure out what the grade was, LSDAS will count both, no matter what the college itself does.

So, the LSDAS folks wanted to use data which colleges couldn’t “fudge” as easily. It’s certainly not fool proof–but it’s the closest thing that’s easily done. LSDAS does NOT have to rely on the willingness of colleges to give them the information.

The LSAT and SAT scales are designed to make it easy to combine them with gpa’s. For the LSAT,drop the 1, divide by 20 and you have a “gpa.” It’s easy to combine with a gpa, so LSs can choose to weight gpa and LSAT equally or weigh the LSAT more heavily or the gpa more heavily. (In the real world, it’s usually the LSAT.) I don’t understand lergnom’s statement about the “relative inflexibility of the gpa.” The LSAT scale correlates perfectly with the 1.0 to 4.0 gpa scale. (It’s assumed that nobody with below a D average graduates from college.) If anything gpa is a bit MORE flexible, since some schools do give A+ grades and it’s not possible to score above an A (180=4.0) on the LSAT.

AS far as I know, no LS CARES whether you “underperformed” or “overperformed” in college. It’s just not part of the calculation, as far as I know. Maybe it’s just yours truly, but I haven’t a CLUE what the last sentence of lergnom’s post means. Please believe that’s not meant as a personal attack–I just really don’t understand what he is saying.

The idea is that in a perfect world each student would receive exactly the same grades no matter which college he attended. While there’s lots of individual variation, the idea is that in the AGGREGATE those who score well on the LSAT should have better grades. So, while there may be a kid like PSedrish’s D who had a 3.2 and a 180 LSAT, it should NOT be the case that the median gpa of all those who score a 180 on the LSAT is a 3.2.

In THEORY, in the AGGREGATE, the 180 test takers should have a gpa of 4.0. The 160 test takers should have a gpa of 3.0. The 150 test takers should have a gpa of 2.5 NO MATTER WHERE THEY WENT TO COLLEGE. This is NOT true of individuals, of course. It’s just that there would be something seriously wrong with the LSAT if as a group those who scored a 150 had worse grades than those who scored a 175.

To get a rough idea of how grade inflated your college is, LSs compare the median gpa and the median LSAT. So, if one school requires everyone to take 2 math classes to graduate and lots of wanna be lawyers don’t do well in math classes, then the overall gpa’s of the kids who attend “mandatory math U.” and apply to LS will be lower than the overall gpa’s of applicants from “Choose Your Own Courses” U. So, comparing the median LSAT with the median gpa of APPLICANTS TO LAW SCHOOL at the two colleges will favor “mandatory math U.” It helps to make up for the fact that the students at “Choose Your Own Courses” U will actually have higher gpas in the AGGREGATE than those at “Mandatory Math U.”

To the extent that X college is grade-inflated at the LOWER end–which seems to be Sakky’s argument–the LSs will know it. The “score report” not only tells LSs your actual LSDAS calculated gpa, and the median gpa, it tells them about where your actual gpa places you among the folks applying to LS. So, if nobody gets below a 3.0 at Harvard–a theory the OP seems to disprove–LSs will know it because the score report for the 3.0 Harvard applicant will say that his/her gpa is in the bottom 10% of all the applicants applying from H.

At the same time, the LSs will know that in our theorectical perfect world, with a median LSAT most years of about 166, the median gpa of H students SHOULD be a 3.3, so it’s not much of a shock that a 3.0 puts our H applicant at the bottom of the stack.

He is no worse off than the applicant from a (real, but I won’t name it) large state U where the median gpa is a 3.1 and the median LSAT is a 149. The median gpa “should be” 2.45. So,the fact that our X state U applicant has a 3.0 isn’t going to look any better than the 3.0 from H, even though in this case, the 3.0 is close to the median gpa at the university.

I don’t know if that helps clarifies things or not. I’m ranting on about this because I get really annoyed when people post things like a 2.8 at Harvard is a “disgrace” or talk about how grade-inflated the Ivies are. Sure, there are colleges that are truly grade deflated–they are rare but there are some. But there’s no reason why H should have the same grade distribution as my large public U example.

jonri · June 1, 2009, 6:25pm

OOPS–time to fix my post has expired, but in the third paragraph I meant to say that there would be something seriously wrong if in the aggregate those who scored 150 on the LSAT had BETTER grades than those who scored 180.

Lergnom · June 1, 2009, 6:53pm

jonri, I believe we just said nearly the exact same thing but that’s obscured because the argument above was that you could use this metric to determine what grades would actually be for the same student at different schools and that some schools are thus easier than others. I can’t see any statistical argument for that from this method of relating LSATs and grades.

To be clear, I’m not saying person x with worse scores is going to be valued more in admissions because he or she relatively outperforms his or her metric-derived expectations. Statistically speaking, one can say x is relatively out-performing versus person y where the metric-derived expectations are higher. You say the same thing: person y would be at the bottom of that school’s pile. Person x might then be at the top of that school’s pile, but how that pile fits a particular law school is not the issue in this discussion.

As a last comment, as you note, they use the information to gather a sense of grade levels at a school and how an applicant from that school fits versus that. To put this in context, the issue specifically for law school admissions is that Harvard and some other schools are treated as having uninflated, even deflated grades. (As I said, it’s not that grades aren’t inflated but that under this metric they’re not inflated, meaning it’s how one views the picture and for what purpose.) As I believe you note, the concern is in whatever subset exists for applicants from a school that has high grades and relatively lower LSATs. I approached this in my head from the perspective of what I believe is the more common worry, where high school students hear rumors of grade deflation and worry that if they went to an easier school, they would get better grades - and thus, with the same LSATs or whatever, would achieve a significant benefit. I think thus we are actually talking about the same worry over inflation, me from the other side.

(One could argue - statistically speaking again - that if you have two schools with similar LSAT scores, you could be penalized if you go to the one with higher grades. The logic is, again, that grades are on a limited scale and that could affect distribution.)

sakky · June 1, 2009, 7:05pm

My point is that there is a big difference between grade inflation as it relates to law school applicants and grade inflation as a whole, particularly as it has to do with tail-end students who don’t even apply to law school, perhaps because they couldn’t even graduate from college at all. The issue is therefore not so much those students at Harvard or anywhere else with the 2.8 GPA, but rather those students with the 1.8, which is below the bare minimum threshold by which you can pass, which for most schools is a 2.0. I am quite confident in saying that - except for cases of academic misconduct - it’s practically impossible to get below a 2.0 at Harvard, but it’s quite possible to do so at a public school such as Berkeley. I knew a number of such people myself. {Heck, I knew one guy whose GPA was literally a 0.5 at one period of time: half D’s, half F’s.}

Those people never apply to law school because they know they won’t get in anywhere. What that means is that law schools will see only a peculiar and highly self-selected subsample of students that is hardly representative of the student body as a whole at any college. More technically speaking, the data is necessarily left-censored.

schiesser · June 1, 2009, 9:21pm

lol @ the long pointless discussions above

anothernycdad · June 2, 2009, 1:13am

I’ve not followed all this closely, but I thought the big deal at Harvard a few years ago was Honors inflation, more than grade inflation - with 70 or 80% graduating with honors, now reduced to 50% or so?

mcat2 · June 2, 2009, 6:13pm

Last time I heard about it, it is 50% at Harvard. At Yale, it is 30% – 5% for summa, the next 10% for magna, and the next 15% for cum laude.

I wonder whether a competive applicant to a T14 law school needs to be in the top 10 or 15 percents of his graduating class, and also whether more students from a particular category of departments (e.g., humanity, social science or physical science) are graduated with summa/magna cum laude. It is rumored that some departments tend to attract competive students.

bluedevilmike · June 2, 2009, 10:02pm

MCAT: I didn’t have any trouble in law school admissions and was not in the top 10%.

bluedevilmike · June 2, 2009, 10:45pm

Well, I was going to tell sakky:*
Sure, I can see some tail-end distribution error, but this is Berkeley we’re talking about. It’s a top-tier university. Its graduation rate has to be what, 95%? It’s not as good as Harvard’s, obviously, but it’s still one of the premier universities in the world.*

sakky probably would have responded:
Look at the data. The dropout rate is higher than 5%.

I would have scoffed.
It can’t possibly be that much higher, can it?

And sakky would have shown me [this</a> chart](<a href=“http://opa.berkeley.edu/analysesandreports/USNWR%202005%20Summary.htm]this”>http://opa.berkeley.edu/analysesandreports/USNWR%202005%20Summary.htm) that Berkeley submitted to US News, and I would have been embarrassed.

With more than one in seven Berkeley undergrads failing to graduate and probably a pathetic less than half managing to graduate [within</a> four years](<a href=“http://opa.berkeley.edu/analysesandreports/Retention&GradRateSummary1.htm]within”>http://opa.berkeley.edu/analysesandreports/Retention&GradRateSummary1.htm), sakky’s point about dropouts may actually be very significant.

Let me think about this for a while.

Thread		Replies	Views
HYP v State School Law School	8	60	April 15, 2021
Law School GPA from Ivy League Undergraduate Law School	5	760	April 15, 2021
GPA bump for attending a prestigious undergrad? Law School	6	74	April 15, 2021
Law School Chances Law School	7	105	April 15, 2021
From Harvard to Harvard Law -- what do I need? Law School	2	61	April 15, 2021

Low GPA from Harvard?

POPULAR STATES

SEARCH SAT SCORES

SEARCH ACT SCORES

SEARCH GPA’S

Low GPA from Harvard?

Related topics

POPULAR STATES

SEARCH SAT SCORES

SEARCH ACT SCORES

SEARCH GPA’S

CONNECT WITH US