The CardBoard
Outside Topic No more grade inflation at Hahvard. - Printable Version

+- The CardBoard (https://thecardboard.org/board)
+-- Forum: C-House! (https://thecardboard.org/board/forum-4.html)
+--- Forum: The CardBoard (https://thecardboard.org/board/forum-5.html)
+--- Thread: Outside Topic No more grade inflation at Hahvard. (/thread-26189.html)


No more grade inflation at Hahvard. - cardcrimson - 05-21-2026

The Harvard faculty voted to greatly reduce the number of A's given to students. Now instead of the 60% university average the last academic year, A's will now be limited to roughly 20% of students per class. That's a h u g e change. 

Interesting to know what the percentage of A's is at Stanford these days. It certainly wasn't 60% in engineering back in the dark ages. . . .

https://www.thecrimson.com/article/2026/5/20/fas-passes-a-grade-cap/

RE: No more grade inflation at Hahvard. - Goose - 05-21-2026

[rant]I do remember a time when the "average" grade was much closer to a 2.0 than a 3.X. Stanford "decided" that since it was so hard to get in to Stanford that by definition everybody admitted was WAY above average and their grades needed to reflect that. One justification for this was that "everybody else is doing it". Of course, the corollary of this determination was that would be redundant to verify these outstanding students actually know the material so it isn't really necessary to verify their mastery by giving hard closed book tests to make them prove it.
 
OK, the above may be considered somewhat of an exaggeration. I am sure there are pockets of  "old school" education at Stanford. It is also true that there are (at least in engineering) lots of mechanical skills that have been greatly revised by the advent of design software. You don't really need a class in mechanical drawing anymore. The more important question is the actual quality of education most graduates possess. When I left in 1973 the consensus was that "Stanford engineers KNOW a lot. However, they don't know how to DO anything." Stanford engineering took up that challenge and started turning out engineers that were much more "job ready". They could actually do many more things right "out of the box". Unfortunately (from hiring experience) what the actually know declined in parallel. IMO that hasn't been a great tradeoff. The angst that CS majors are feeling about AI reflects this reality. Too many of them see themselves as coders, not  actually "Computer Scientists". Unfortunately many of them are correct. [/rant]

RE: No more grade inflation at Hahvard. - 82lsju - 05-21-2026

(05-21-2026, 07:01 AM)cardcrimson Wrote:  The Harvard faculty voted to greatly reduce the number of A's given to students. Now instead of the 60% university average the last academic year, A's will now be limited to roughly 20% of students per class. That's a h u g e change. 

Interesting to know what the percentage of A's is at Stanford these days. It certainly wasn't 60% in engineering back in the dark ages. . . .

https://www.thecrimson.com/article/2026/5/20/fas-passes-a-grade-cap/

big change in large classes maybe not so much in small ones.  They can give up to 20% As with up to four additional As per class so in a class of 100 that's up to 24 (20% +4) As, so up to 24%, in a small class of say 15 that's up to 7 As (20% +4) or 47%.

last fall they apparently tightened up grading a bit (but not enough)

Quote:Faculty have already taken steps to curb grade inflation, slashing the number of As awarded from 60.2 percent to 53.4 percent last fall. But the committee argued that voluntary reductions were insufficient to preserve the A as a mark of “extraordinary distinction.”

https://www.thecrimson.com/article/2026/2/6/faculty-propose-grade-cap/

RE: No more grade inflation at Hahvard. - jacket3ree - 05-21-2026

It's been so long, but my memory as a TA in my first quarter of teaching a lab at Stanford and giving a (gasp!) "C" on one of many assignments to come that fall to apparently an entitled complainer does not mesh with many of the stories of academic brutality I've seen on the Board recently.  I was called to the carpet for that C, which as I explained to the professor who owned the carpet, was frankly it was crap work and I thought I was being a nice guy. Prof's response was that while I may have come from Hardass U (or I, I suppose), these are Stanford students and they don't do crap work. Prof was on his way out and I think what he really meant to say, without saying it, was 'take care of this so I don't have grade-grubbing students in my office while I watch the approaching sunset.' My solution was to be extremely helpful marking up lab reports and tests to emphasize what they were suppose to be learning and worry less about the letter grades. It turns out, this was great for me! That year of being the TA for E21 (fluid mechanics) probably taught me more about hydraulics, pressure flow, open channel flow, groundwater flow and the rest than anything else. I had to make the test and homework keys and assign the relative points to each solution, and that taught me a lot too. Major points for getting the concept and approach right. Minor dings for messing up some mundane detail like putting a decimal point in the wrong place or something. I mean corporate is sure as hell gonna notice $305,326.13 but as long as they know Bernoulli and when to use that instead of momentum; all good!

Anyway this technique got me a free lunch at the end of the year as a reward for (apparently really) good student reviews.

I do recall some prospective employer once calling about a Stanford student who had worked for me before going back to grad school. "Oh, I see he has almost a 4.0 but I suppose they all do there."

All anyone wants is fair. Work hard and do well - be rewarded. I recall getting a 7 (out of 100) on an emag test grinding through the required sequence in Physics. That was a solid C but I was not learning material because good old Rudolph "Hess" Arrans was having so much fun tricking us all and taking great delight in telling us we are all stupid and these scores prove it. Sheesh. Good lessons in perseverance and learning to plow though tricky clients and Councils, but not helping me understand a damn thing about electromagnetism. I'm more a fan of objective scores that reward mastery of material (and/or great memorization - whatever it takes) than forced curve setting and Lord of the Flies culture.

Stanford's current undergraduate GPA is 3.65, FWIW.  Some employers obsess about that while many don't. I guess law schools and med schools might too. My son's graduate GPA was about 0.7 point higher than what he pulled at USNA. I only know because he told me both times, nonchalantly. I'll always remain more impressed that he finished first in his class at The Basic School. And if I'm doling out pride, my daughter was basically a straight A student. More importantly, they are both great people. Kudos to Mrs. Jacketree on that.

Grad school is different (calling professors by their first names!) and I have no idea what I did at Stanford because I decided never to open those envelopes and no one ever asked. Still haven't seen anyone's GPA or SAT score on a headstone nor heard them raised in any eulogy.

The blowback from Harvard students will be interesting.

One Stanford student's recent opinion on this topic

RE: No more grade inflation at Hahvard. - Nan3cy - 05-21-2026

Wow. Back in 1969 a 3.7 could get you Phi Beta Kappa.

RE: No more grade inflation at Hahvard. - needle - 05-21-2026

I was a mostly B's, more A's than C's kinda student at Stanford.

That said, I'll take advantage of my (relative) anonymity to to say what I think is way more interesting: I cheated.

Not every time I could get away with it, but when I saw an answer key left out before the problem set was due, I swiped it. If the problem set solutions from the previous quarter were left online, I downloaded them all. My memory doesn't provide me with any other examples, though there might have been a couple. That said, I honestly would have *never* cheated on a midterm / test. FWIW.

I hear cheating via AI is much more rampant nowadays, so I wonder if that somehow factors into what Harvard is doing to adjust its grades and what perhaps Stanford is considering doing. I have just enough teaching experience to know that many young people take pride in cheating (and getting away with it), which is sad. I always felt shame when I did it, though not enough, obviously.

RE: No more grade inflation at Hahvard. - Jamesy - 05-21-2026

(05-21-2026, 08:33 PM)needle Wrote:  I was a mostly B's, more A's than C's kinda student at Stanford.

That said, I'll take advantage of my (relative) anonymity to to say what I think is way more interesting: I cheated.

Not every time I could get away with it, but when I saw an answer key left out before the problem set was due, I swiped it. If the problem set solutions from the previous quarter were left online, I downloaded them all. My memory doesn't provide me with any other examples, though there might have been a couple. That said, I honestly would have *never* cheated on a midterm / test. FWIW.

I hear cheating via AI is much more rampant nowadays, so I wonder if that somehow factors into what Harvard is doing to adjust its grades and what perhaps Stanford is considering doing. I have just enough teaching experience to know that many young people take pride in cheating (and getting away with it), which is sad. I always felt shame when I did it, though not enough, obviously.

I'll use relative anonymity to get something off my chest. During an exam, I saw two students blatantly cheating. They were sitting with a buffer seat between them. One would be working in one of the blue answer books, put it on the buffer seat desk, then the other student would pick up that book and either copy the answers down or switch it with one of their own. I knew both of the people involved and still think about the incident whenever I hear about or see their info online. I didn't want to snitch but I regret not doing so.

RE: No more grade inflation at Hahvard. - TonyLima - 05-21-2026

Fun stories. 

My first year at MIT was 1965. Completely C centered curve. One or two of the guys in my pledge class flunked out freshman year. Mind you, they had been good enough to be admitted to the place. Fast forward, I’m teaching econ at Stanford (I forget the exact class but it wasn’t principles). After the quarter ended, a student came into my office asking me to raise his grade from A to A+.  In that class of about 130 students I gave one A+ to a student who was way better than anyone else. I believe I told my office visitor that he was on his way to a C if he didn’t leave quickly. 

Fun times.

RE: No more grade inflation at Hahvard. - Mick - 05-22-2026

At SCU, they published the overall student body GPA at 2.68 after my frosh year in 1980/1981. I distinctly remember thinking that if I'd known what the average GPA was in 1979/80, I wouldn't have even applied to the stupid place...

My oldest son was at Princeton during their period of official, formal, curve-forced grade deflation (2000-2014). Apparently Princeton stopped when (a) the other Ivies refused to match, and (b) employers rejected the "3.2-at-Princton-is-the-same-as-3.8-at-Yale" trope. It cost Princetonians numerous opportunities. Only 5% of students and 6% of faculty thought it was effective. Last year, 2/3rds of Princeton grades were A-, A, or A+.

RE: No more grade inflation at Hahvard. - jacket3ree - 05-22-2026

Needle,

A lot of what you are talking about is what we used to call "Word". I remember going to frat parties when I first landed in ATL (I guess I rushed without really trying) and we all got to see the Word files/cabinets/shelves.  Studying prior year exams is not cheating, it was excellent preparation. If professors are that lazy to repeat questions year after year, that's on them. At Stanford, one of my friends (she earned a PhD and became a tenured professor at a prestigious university) was caught red-handed with a prior exam during an open-book final exam. Interesting exchange that we all witnessed. "That looks really familiar!" She didn't flinch. "It should be - it's your final from last year." "Did you steal that?" "No, we all have friends, you know. This friend took this same class last year. If you are too lazy to make up new exam questions, that's on you, not me!" Perfect umbrage. "OK, fair point." Afterwards I asked what she was thinking when sternly talking him down. Pretty bold, L!. 'I was thinking don't throw up.' 

Thirty years later, we tore down our decrepit house (but it was our decrepit house) and built a fabulous new one. New houses are built off the backs of alumni brothers who give money and get rooms named after them. I'm proud to say that my room is the Word room (aka library). brobdingnagian upgrade from the filing cabinet at the end of the hall with all the Word in it, and the shared vinyl.

Went to the Darden (UVA) graduation last weekend. Everyone mentioned AI and that their experience should help them survive the coming robot apocalypse.  One of the better speeches was clearly AI written and as everyone started to murmur, the grad shredded some paper and 'took off the shackles of our robot overlords and set [himself] free!' then talked a lot about how everyone needs to find a purpose, even if that purpose is just ordering great drinks for their improv club.

And if everyone gets an A, no one does.

RE: No more grade inflation at Hahvard. - cardcrimson - 05-22-2026

(05-22-2026, 08:24 AM)jacket3ree Wrote:  Needle,

A lot of what you are talking about is what we used to call "Word". I remember going to frat parties when I first landed in ATL (I guess I rushed without really trying) and we all got to see the Word files/cabinets/shelves.  Studying prior year exams is not cheating, it was excellent preparation. If professors are that lazy to repeat questions year after year, that's on them. At Stanford, one of my friends (she earned a PhD and became a tenured professor at a prestigious university) was caught red-handed with a prior exam during an open-book final exam. Interesting exchange that we all witnessed. "That looks really familiar!" She didn't flinch. "It should be - it's your final from last year." "Did you steal that?" "No, we all have friends, you know. This friend took this same class last year. If you are too lazy to make up new exam questions, that's on you, not me!" Perfect umbrage. "OK, fair point." Afterwards I asked what she was thinking when sternly talking him down. Pretty bold, L!. 'I was thinking don't throw up.' 

Thirty years later, we tore down our decrepit house (but it was our decrepit house) and built a fabulous new one. New houses are built off the backs of alumni brothers who give money and get rooms named after them. I'm proud to say that my room is the Word room (aka library). brobdingnagian upgrade from the filing cabinet at the end of the hall with all the Word in it, and the shared vinyl.

Went to the Darden (UVA) graduation last weekend. Everyone mentioned AI and that their experience should help them survive the coming robot apocalypse.  One of the better speeches was clearly AI written and as everyone started to murmur, the grad shredded some paper and 'took off the shackles of our robot overlords and set [himself] free!' then talked a lot about how everyone needs to find a purpose, even if that purpose is just ordering great drinks for their improv club.

And if everyone gets an A, no one does.

Thanks for the post. The old engineering lab in Terman had old exams and answer sheets on file that you could check out and study. In some cases, five or more years of exams. Again, excellent preparation. And yes, professors did recycle questions. . . .

Yes. If everyone gets an A, no one does. Very reminiscent of a poster who signed his posts with "orange slices for all". . . .

RE: No more grade inflation at Hahvard. - BostonCard - 05-22-2026

I saw this one online:

https://www.sfchronicle.com/politics/article/uc-berkeley-law-school-ai-22271280.php

Berkeley Law School bans AI use because of plagiarism.

Quote:In a policy announced Thursday, and effective this summer, the law school will prohibit its 1,120 students from using artificial intelligence in preparing class assignments and in all exams — except for courses designed to teach students how to use AI, ethically and legally, while practicing law. Those courses are generally for students who are nearing graduation.

Exceptions will also be made for teachers who want to give their students more leeway in using the technology. But it will generally be barred for end-of-term exams.

Students will also be prohibited, the university announced, from using AI devices “for aid in conceptualizing, outlining, drafting, revising, translating, or editing any work submitted for credit.”

This strikes me as very short-sighted.  There is no stuffing the genie back into the bottle, so instead of prohibiting it (after all, I presume practicing lawyers are using AI, though sometimes getting into trouble through its mis-use), you might as well incorporate it into law school.  That would require changing the way you teach, and possibly how you evaluate students.  And if there is an issue with plagiarism or hallucinations coming from the AI, then the school has to teach students how to spot it, or check for it.

At my current job, we don’t get told not to use AI (though we do have lists of approved and prohibited tools); rather we have had mandatory everydayAI training courses designed to help people incorporate AI into their workflows, and figure out how to use it, how to check it, and (also critically) when not to use it.  I suspect my company is not unique.  That’s going to be the future of new grads coming into the workforce, and as much ink as has been spilled on how AI will do away with entry level jobs, I actually think it is the opposite; new grads coming into the workforce will be the first AI “native” generation, so they will be able to incorporate AI into their workflows seamlessly, and much more effectively than fuddy duddies like me.


BC

RE: No more grade inflation at Hahvard. - jacket3ree - 05-22-2026

We could almost do a special thread on AI.  We use it if clients allow (some don't) and disclose its use no matter what, other than glorified searches on steroids. Two recent anecdotes:

During a deposition the opposing attorney was running through my file. Some obvious AI generated content. "What's this?" "A summary of the thousands of pages of Bates stamped transcripts when I was looking for keywords x, y, and z." "Is this foundational to any of your opinions?" "No." "Then why is it here?" "To save me many hours of reading transcripts looking for something it turns out isn't there." "Cool, then I don't give a shit." (On transcript!)  Not AI related, but the attorney I'm working with told me the other guy is a lying sack of ..... sugar. So when she turned it over to another guy for the defense, but a different defendant for questions, he said "Would it shock you if what he just asked you was [full of shit]" about 8 times in a row for each of the FOS questions. Hilarious.

My wife's birthday, the last non-round number birthday for a year is coming up and I searched for a getaway destination. Helpful narrowing it down but I did the work and booking. Then why not? "Recommend a route that avoids freeways and is suitable for a 240hp convertible."  AI sucked at that. I know great routes and every 'why are you backtracking me' query was met with 'oh you are exactly right, this is perfect now.'  Or as Ms. Jacketree will find, we are just headed north and headed east. If it worked in Italy when she was so petrified/angry (petrangy) that I spent the night on the floor, it will work where I actually do know the backroads. My schtick is to not tell her where we are going and make sure to drive through Los Banos to scare the crap out of her before it is obvious we're headed to Yosemite, etc.  "What do I even pack?" "Clothes. Or.....nothing; I mean it's your birthday." Last year's birthday epic was indeed Yosemite through Hollister ("Monterey, right?" Nope.) and Santa Barbara on the same trip via Maricopa and Ojai and yes that was pushing it to drag her through 100 degree heat in the Valley on CA 33 with an open air car. Frowny birthday face until we hit the coastal fog.

RE: No more grade inflation at Hahvard. - Mick - 05-22-2026

Last month, a large NYC law firm (nearly 1,000 lawyers) filed a major motion in a high profile case that required forty corrections for AI-related errors; hallucinations, 

Irony of ironies, this firm advises OpenAI on safe and ethical deployment of artificial intelligence. Second irony: the partner who had to write the apology/please-don't-sanction-us-for-all-time letter also represented Silicon Valley Bank as debtor.

Just for fun, a database cataloging nearly 1,500 AI-related errors in litigation:  AI Hallucination Cases Database – Damien Charlotin

Last item: I am publishing my third book next spring, and this was the first time that the publisher not only asked for my AI use for any part of the book (none at all) but also used AI tools to check the manuscript. My editor told me it was for two reasons, both fairly obvious: plagiarism and the fact that AI writing just isn't that good, at least not yet.

RE: No more grade inflation at Hahvard. - BostonCard - 05-24-2026

A lesser publicized reform as part of the effort to combat grade inflation is that honors and other distinctions are no longer going to be based on GPA, but rather on "average percentile ranking".

https://current.fas.harvard.edu/stories/faculty-decisively-approve-grading-changes

BC

RE: No more grade inflation at Hahvard. - 82lsju - 05-24-2026

(05-24-2026, 07:46 PM)BostonCard Wrote:  A lesser publicized reform as part of the effort to combat grade inflation is that honors and other distinctions are no longer going to be based on GPA, but rather on "average percentile ranking".

https://current.fas.harvard.edu/stories/faculty-decisively-approve-grading-changes

BC

I wonder how different average percentile ranking will be from GPA ranking for the awards.  I also wonder if this will cause students to select smaller class as the percentage of As could be much higher in small classes given up to 20% +4 As is potentially a lot higher in a class of 10 than a class of 100.