Are Easy A’s Too Widespread in Academia?

Well, maybe it's worth just talking through the logistics of what Scott's proposing.

You know, he's, he's pr- maybe he's proposing something like an adjusted GPA that could be printed on your transcript.

You're essentially, you're essentially estimating a fixed effect for every class-

Mm

... more or less.

And so you're adjusting all of the GPAs accordingly by class, depending on both, you know, how hard the class is and how good the students are in that class.

And so because you have a bunch of different students all taking a lot of the same classes, you have the same student in a hard class and an easy class.

You can kind of tease out mechanically, this is a hard class, this is an easy class, and then you essentially partial all that out.

You could run one big regression.

Yeah.

And you end up getting an adjusted GPA for every student.

And the proposal is something like that, that imagine that was printed on the transcript, and you could say, "This student got a three point nine at Harvard, but their adjusted GPA is a three point three 'cause they took mostly easy classes."

Doing this solves two problems.

The problem number one is I see someone who got an A in one class and someone who got a B in another class, and I don't know whether the first person is better or maybe the first class was harder.

So now we have a way of grading where if I see someone with a higher grade, I know the student is of higher ability.

So that's problem number one, yes?

At the GPA level.

At the GPA level.

Problem number two, it should solve selection, and that's like, I guess, a question to you.

Because now if you take an easy course, your grade will be adjusted downwards.

So all else equal, you don't have any preference across the difficulty of the courses as to which one to take.

Your ability will be revealed no matter which one you take.

Is that correct?

If you take enough-

Yes