@efle @GrapesofRath
Here is what I can say about my school. I go to an Ivy League school, and here is my experience. This is only directly applicable to my school. I do not think my school has a reputation for grade inflation or deflation other than "zomg its ivy therefore its super inflated
"
All of our hard sciences classes (with one exception) were curved to a B median. However, our distributions on these curves were not equivalent for all grades. Our medians were B range as well as our mode. Generally, it looked something like this:
15-20% A/A-
60-80% B+/B/B-
5-15% C+/C/C-
0-5% D/F
It varied depending on the class, but it was generally this. I had one biology class where only 3 people got As and 2 got A-s, but the median was a B. I had another biology class where anyone with a 90%+ raw score was automatically given an A while anyone else was subjected to the curve scaling (so you might get an A with below a 90%, but it's not guaranteed). The median was again a B. I had a third biology class where generally exam scores had medians in the 70s and you were graded based on a curve with the 50% percentile being a B, but the standard deviation on raw scores being very small, so in order to hit the A threshold, you had to score significantly above the median. This was the case with my chemistry classes too, though the median (raw) grades and ranges in chemistry were usually lower than in biology classes by 5-20 points. One chemistry class I took had a B- median about 1/3 times it was offered (when I took it, it was a B median, though friends and underclassmen took it with the B-).
The distribution I presented above is a general trend based off of the histograms that my professors gave to us at the end of each test along with the cutoffs for each grade (they generally gave us A B C D cutoffs without the +s and -s except in certain circumstances or if you asked them where you stood because you were on the border).
Most of our grades came from exams (3-4 total depending on the class) with lab integrated into our class and essentially counting as another exam. For instance, if a class had 3 exams, it would generally be 25% midterm 1, 25% midterm 2, 30% final, 20% lab, or something similar to that. In some cases, 1-2% would be "clicker" questions, which nearly everyone got because at my school, nearly everyone comes to class all the time (even when these clicker questions don't exist). Our exams are all made by our professors and come from their lectures, notes they write on the blackboards, and powerpoints (in classes that had them). In my experience, all my bio classes had powerpoints while my chem and physics classes were all written on the blackboard (so if you missed a class you had to get notes from a friend), though things were also written on the board in most of my bio classes in addition to the powerpoints, which made note-taking vital.
Now, with all this said, we see that you had to score in the top two deciles in pre-med heavy classes at one of the best schools in the country that is packed to the brim with ambitious, bright, motivated, and disciplined students. These students were all gunning for As, all (if you feel this is a valuable metric) scored in the top percentiles of the SAT/ACT, all invested a ton of effort in these classes in addition to everything else they needed to do, and yet still you had to outcompete at the very least 80% of them to even get an A-. Thus, in order to match the national median for accepted students to MD schools, you had to consistently be outperforming 80% of equally talented premed students whose big claim to fame is that they are good at schools. This is far more difficult than some people might expect. Many of these students, at a school with less competition, would be pulling As in these classes, but instead they are getting Bs because 20+% of the premed body performs better than them on these tests.
Now this is only for Biology, Physics, Chemistry. Other departments (including Neuroscience, which has a reputation within my school [or maybe just in my head] for being science for premeds who can't hack it in biology) have much higher department GPA medians. My school put out a departmental report based on overall median for that department and the range was astounding. You had nearly a 0.8 difference between the lowest 5 (which included biology, physics, and chemistry) and the highest 5 (which included stuff like theater and other humanities).
I absolutely do not deny that my school is inflated in the humanities and parts of the social sciences (ie everything except our econ department which is also in our bottom 5 and the psych department). Medians in many many many humanities classes are A or A- (with some B+s thrown in there) and I think it's ridiculous. This is why you might commonly see non-science major premeds with cGPAs of 3.7 but sGPAs of 3.2-3.3.
Now, yes, I understand schools like WashU have similar levels of competition, rigor, etc, and yet have C median classes, but I posit that the percentage of students getting As and A-s (ie what really matters when you're applying to medical school) in both cases are similar (somewhere between 10 and 20%), so if you're still at the top of your class, it won't matter either way, but it can hurt more at WashU than at my school if you're in the middle of the pack.
As with WashU, our school's average successful MD stats are different than the national median. The national median is 3.7/31 while for our school, it is around 3.55/33-34 (these numbers are nearly 2 years old now though, as that's the last time I talked to my premed office about this stuff, so it may have changed slightly).
Our tests in the sciences emphasize critical thinking over memorization, and although you won't get too far if you don't know things, you are certainly not going to do very well on our tests. Multiple choice tests? What are those again? Most of my core premed classes had exams with 6-7 questions and 3 hours to complete them in. No (or very little) multiple choice. It was all about using information to figure things out and piece together bits of knowledge until you can come up with a cohesive solution. And although this can be very difficult, it makes the MCAT (at least old MCAT - I have no experience with the new MCAT) science sections essentially a joke. People with 3.4 GPAs would blast the MCAT out of the water without exhaustive studying (again, I'm sure this trend exists across all top schools). I don't think my roommate's science GPA even hit a 3.2, but he did exceptionally on the MCAT (which he took in the middle of a heavy term - the day after a midterm no less) and he got into a top 30 med school (didn't have anything else fancy either). We have an MD acceptance rate upwards of 80% (won't say the exact number) and we absolutely do not screen our applicants in any way (I've been through this process so I know).
So the short answer to "are Ivies inflated" is yes but it's more like yes* because it varies across department and a B median doesn't mean that you're any better off applying to medical school than you would be at a school with C medians instead.
If you have questions about my experience or my response, I'm happy to answer them. I think I explained this fairly well, so I am going to link to this post in my signature and will point people towards it whenever something related to this issue comes up.