PDA

View Full Version : What do we mean when we say grade inflation?



Econhead
11-19-2014, 10:35 PM
Grade inflation is thrown around a lot. I am curious-What exactly is everyone referring to when they say that it is a "problem."

From my experiences, there are several ways that this could come up:

1. Professor thinks that you are in "the big leagues," and essentially breaks the grades into a distribution of A and B (with some distribution of plus's and minus's).
2. Professor is extremely difficult grader (i.e. high is ~50-60, avg. ~20-30), and as a result a huge 'curve' is attached.
3. Professor has low expectations, and as such grades are abnormally high.
4. Professor grades in such a way that they want approximately X% to be A/B, resulting in disproportionately many in these categories (and few below).
5. Professor has an expectation in mind of where she wants students to be. If they meet that, they are in the A/B range (and almost all students meet that); regardless, she does not try to make it difficult enough that students actually fall into some type of A/B/C/D category, but leaves it as A/B.

I attended an unknown state school for my undergrad with no graduate econ program. I've experienced all of these. I view all of them except #2 (and to some degree #5, based on circumstances) as being a problem, yet #2 is the one that I have encountered most frequently (almost exclusively in math courses). In fact, I think that I have only encountered the others only perhaps once or twice each. To me, the problem with 1, 3, 4, and 5 is that the professors don't use the entire grading scale - they are 'against' giving lower grades, regardless of why.

What has everyone else encountered? Am I in the minority to have primarily experienced #2 (which seems less of a problem)? Is this what everyone refers to as "grade inflation"?

Blanket
11-20-2014, 12:26 AM
4 is a big problem. I transferred from a directional state University without a Ph.D program and experienced none of these, the material was (relatively) easy. I transferred to an elite University and am experiencing, and frustrated by, #4. I work my tail off to get 20 points above the mean on our first midterm, a mean of 62, and a ton of those around the mean are going to get, or have a shot at, an A. Pretty frustrating stuff in the sense that an admission committee sees identical work for every A in the class.

Econhead
11-20-2014, 03:55 AM
4 is a big problem. I transferred from a directional state University without a Ph.D program and experienced none of these, the material was (relatively) easy. I transferred to an elite University and am experiencing, and frustrated by, #4. I work my tail off to get 20 points above the mean on our first midterm, a mean of 62, and a ton of those around the mean are going to get, or have a shot at, an A. Pretty frustrating stuff in the sense that an admission committee sees identical work for every A in the class.

I'd say that you have just identified one of the key reasons for why letters of recommendation are often given so much weight.

Blanket
11-20-2014, 03:25 PM
I'd say that you have just identified one of the key reasons for why letters of recommendation are often given so much weight.

Now if only my Professor spoke English and the class had less than 80 students...

Econhead
11-20-2014, 03:32 PM
Now if only my Professor spoke English and the class had less than 80 students...

I didn't necessarily mean for any single course, but rather speaking to a candidate's skills holistically. That said, if you're not top in the course or one of the top students in the course, I can't imagine it'd be that great of a letter without having spent significant amounts of time speaking to your professor on the side.

All letters should be positive. Amazing letters should tease out the nuances of what makes student X amazing - what separates someone who got an A in game theory from someone who got the highest grade, had the best gasp of the material, or whom apparently has intuition for days. Grades can be read from transcripts - these nuances should be what can't be garnered or deciphered from transcripts. Along that same line, suppose you were one of 3 students in a class of 80 to get an A. The transcript also will not show that you did so well compared to your peers, only that you received an A. One would expect that if you did that well that the professor would have other things to say as well.

Mathew952
11-21-2014, 02:18 PM
I think within-variation is actually a bigger problem. For example, I am in a super intense Honors Calc III course right now, but It's curved to the point that I could basically kill the engines right now and get an A. But then I'm in a probability theory course that's barely harder than the fakey stat class they make the business students take, but the entire grade is determined by exams that he grades with absolutely no mercy (I lost two letter grades because I summed one series wrong on one part of one problem). Not only does this totally miss my actual skill levels, I think it makes grades have a lot more noise than signal. If I could bomb my Calc final and still get an A, but come close to perfect on a prob theory exam and get a 78 over two minor mistakes, it really seems like the grade is a terrible predictor of anything.

Blanket
11-21-2014, 09:01 PM
I think within-variation is actually a bigger problem. For example, I am in a super intense Honors Calc III course right now, but It's curved to the point that I could basically kill the engines right now and get an A. But then I'm in a probability theory course that's barely harder than the fakey stat class they make the business students take, but the entire grade is determined by exams that he grades with absolutely no mercy (I lost two letter grades because I summed one series wrong on one part of one problem). Not only does this totally miss my actual skill levels, I think it makes grades have a lot more noise than signal. If I could bomb my Calc final and still get an A, but come close to perfect on a prob theory exam and get a 78 over two minor mistakes, it really seems like the grade is a terrible predictor of anything.

I agree. The grade is a terrible predictor. Not only that, but attaching a grade to one's performance incentivizes getting a good grade, not necessarily engaging with the material in a meaningful way. Yesterday I sat through an Intermediate Micro exam with more rampant cheating than anything I had ever experienced before. It was absurd. I literally saw a student stand up to peer over a girl's shoulder standing in front of him. Two international students next to me were having a hushed conversation during the test.

Evergreen State University does not have grades. Instead, their Professors assess each of their students' performance in the form of an essay at the end of the semester/term. I like their system best, but it has its obvious flaws, both in its subjective nature of assessment and practical concerns for the University ($$$).

ColonelForbin
11-21-2014, 09:38 PM
Reed College gives grades, but you only get to see them AFTER you graduate.

Now this only works for an extremely special student body (IMO), but I like the idea.