Grading on the curve

Chris's posted some continuing thoughts on the question of whether its a good idea to grade people on a curve when doing performance reviews (or any other place where grading occurs, I guess). Like Chris, I also believe pretty strongly in grading on the curve but I still have lots of reservations about it. That's because a curve can so easily be misused in a way that is damaging to employees, the company, or both. But the alternative, inevitable grade inflation, seems to be worse.

(Grade inflation is something I got enough exposure to in college, thank you very much. Which is not to say that I wasn't hypocritically disappointed that my university started giving out graduation honors on a curve starting with my class. It turned out something like over 50% of the class was graduating “cum laude“ or better. It is numerically impossible that every student at a university is above average.)

Ultimately, it reminds me of the famous Churchill quote: “Democracy is a very bad form of government. Unfortunately, all the others are so much worse.” Same for the curve, same for the curve...

Feedback

# Curves

I've been following a loosely coupled weblog discussion about using curves when grading, doing performance reviews, etc. It's being discussed by Paul Vick here. Chris Anderson talks about it here and here. It all started from John Porcaro's entry here.... 9/26/2003 9:00 AM | Hurricane Blog

# re: Grading on the curve

I offer these possible grading curves:

www.alexander-thegreatgradingscale.com

Thanks 2/3/2006 8:44 PM | RW Alexander

# Grading on the curve

I recall back in antiquity a neat statistical program for the computer which will assist in the process of "grading on the curve". For the life of me, I don't seem to be able to find such a program for downloading from the Internet WEB.

Perhaps you can point me in the right direction.

Certainly you know what I'm looking for - namely, a program into which one may "plug in" a population of raw grade scores - then, after having specified the desired median score and standard deviation, such program will yield a new spectrum of adjusted grades scattered on the normal distribution curve.

Maybe that little program is hidden in one of the "stat pacs" floating around.

Gratefully,

David Watters
Franklin, NC
EE Ga Tech '50
thornbush@dnet.net

6/21/2006 10:55 AM | David Watters

# re: Grading on the curve

I don't believe that grading on a curve is fair. Each individual should be responsible for passing or failing through his or her own efforts. I'm taking a statistics class right now (beginning) and from what I understand about curves and grading, it wouldn't matter how hard I studied or how well I did if a certain percentage of students beat me to the curve. I believe that if an 90 represents and A, then it should be an A for all students who take the test. And what if there is a "tie grade" and the "20% in the A category" is filled before the 2nd grade gets put in? Who gets an A and who gets a B+? Most likely the white man. 7/1/2007 5:11 AM | Monica

# re: Grading on the Curve

Thank you Muzzy - That is exactly what I wanted.

And for you, Monica, a pox on your 'anal retentiveness' and racism.

The design of a fair quiz examination is largely dependent on the teacher/professor's skill. When that skill is outrageous - packed far above the expectations of the brightest students in the class - then, in fairness, some grade adjustment is called for.

Ideally, the median scores of a class should be about 75 or 80. If that mean is much lower, that brings into question the teaching qualities of the teacher/professor or his/her quiz examination design. This rational holds for a population of bright students and for a classroom of dummies. John Dewey was not always wrong in his theories of education.
10/4/2008 9:26 PM | David Watters

# re: Grading on the curve

@ David Watters

which is it... "median scores" or "that mean?" 2/27/2009 1:33 AM | Your Mother

Panopticon Central