I agree to some extent, Edwin-- but as my DH quickly noted when I mentioned this phenomenon--
statistically, it seems somewhat unlikely that MOST students attending even an "elite" institution have so little to learn that they are earning A's at everything.
I agree with him. While I loathe the application of normal distributions in determining grades in a college cohort (well, any cohort, really, because I subscribe to mastery-oriented benchmarking there), I also know from experience that students TEND to distribute themselves more or less into a normal distribution in introductory coursework, and a top-heavy version of it later on in upper-division courses.
I've never taught an undergraduate class where 90% or more of students earned some kind of A, and my grading always allowed for it. Now, one can certainly argue that the quality of students I was seeing was significantly different from the average at Harvard, and that's probably true, but still-- the students from there that I've rubbed elbows with at meetings sure didn't come across as THAT much different.
The highest percentage I ever personally saw was 14 A/A- grades in a class with 32 students in it. I thought that was very high-- but what the heck, they had mastered what I set for them, so there it was. KWIM? Even so, the average that term was still an 82%. There were just a lot of really great students in that particular section, which overlaid the regular distribution. But that regular distribution was most certainly still PRESENT.
So I'd say that it may beg the question of either methodology/integrity of assessment or, as Edwin noted, the level of the curricular offering.
I know someone who was TOLD to reevaluate a final exam in an undergraduate chemistry course at an Ivy whilst a graduate teaching assistant there. The student in question was the daughter of a former Atty. General. The college dean very clearly told the grad assistant that "there has clearly been a misunderstanding" regarding the undergraduate's (well-deserved) low C grade on that exam. Apparently, it would need to be "looked at again" unless it were an A.
That was in the mid-90's, by the way.
