Log in

No account? Create an account

Previous Entry | Next Entry

Emergent questions about grades


Grades are in.

Why is it that there are more thirteenth-hour questions than there were eleventh-hour ones?

I mean, I know I try to put students at ease when they needn't be alarmed, just because some of them are very antsy, and some are just a little insecure. Some students should be worried, though! To wit: grad students on the brink of a C or undergrads on the brink of a D or F should hit the books (or come and check on missing homeworks or their absolute standing) before the final. It's easy to say "I didn't see a grade posting, so I just guessed (read: assumed) I was okay"; it's quite another to know you only turned in half the assignments or turned the hour exams in half blank and then count on the curve.

... right?



May. 22nd, 2006 06:59 am (UTC)
"Large" curve...well that depends on context.

ECON 110 (Intro to Macroeconomics) would not just annoy you. The semester I took it (at KSU):

A: 6?/100 for all hourly exams, scale appropriately for 200-pt final.

And this is for a class that can be handled virtually math-free as a foreign language. I understand the desire to not C-out every single business major, but give me a break: properly taught, business is not a content-free major. The number of specializations required to correctly expand a startup beyond a sole proprietorship is more than one person can conveniently learn.
May. 22nd, 2006 07:10 am (UTC)
Curve does depend on context, if a "large" curve is needed, because either the class wasn't taught properly, or for another reason, in other words, a large curve is justified then so be it.

The class I'm irked with had 29% of the students who got A's without any curve, the class average was a 78.7%. Most professors I know will give a curve if the average is below a 75 or a 70%.

If the class average is a 53% (like a class my Math major friend took last year), then I say 1)something is wrong with the class and 2)a curve is warrented, but not at a 78.7%
May. 22nd, 2006 07:12 am (UTC)
I should also add, I normally don't have a problem with a huge curve, I've benefited (happily) from them before. It's when a student does a crapload of extra credit work at the end of the semester AND doing substandard work during the semester (and not going to get help) AND fails the final that I have a problem with them getting an A.

If you're really trying, hey, that's a whole 'nother story.
May. 22nd, 2006 08:22 am (UTC)
Class average 53%...I've seen worse in KSU MATH. But there are two or three professors I respect (e.g., Saeki) that could do that in a mid-level course without doing anything wrong. (They just grade hard...very good for learning what a proof is. Saeki crushed my first semester of grad school GPA, but it was worth it.)

In low-level, I saw was a College Algebra with a 65% average...after dropping the one-third of scores less than 40/100. I felt this addendum deserved elevation from fine print. (Got rid of 1/3rd of my section...but really, there was no point leading them on a charade.)
May. 22nd, 2006 03:42 pm (UTC)
Grade inflation and dropping flunkers
At UIUC there is a feature of Rob Hasker's famous gradebook program gr, a version of which you can still find, with source code IIRC, on the net, that does "min > 0". This drops the no-shows on homeworks and midterms, even if they are still in the course.

In general, I think this is meaningful, and dropping 10%-ers and 30%-ers is less so.

My dad favored the "square root times 10" curve at National Taiwan University - I've never used it, but there is some sense to it, too.

Bottom line: a nonlinear scaling of grades is OK, as is lowering the cutoffs for A/B/C/D to 850/700/550/400 as I do. It's when you can pass, or earn a C, for having done no work, and just being qualified to come in the door, that I object.

May. 22nd, 2006 06:33 pm (UTC)
Context is everything, and getting there is half the fun
(Comment title from a quotation by J. M. Straczynski.)

"Math-free as a foreign language"? I'm not sure I get that.

Is "C-out" an analogue of "flunk out"? If so, why? I understand that D's are now the "new F's" because of probationary ramifications of C's and D's. (They shouldn't be, IMO; F's are F's, and ramifications are there for a reason: to create a buffer for people and to give them ample warning.)

As for 60's on all exams earning an A: on the one hand, knowing how it is at a few other universities, I agree that this is a crock. On the other hand, if the whole point scale was being utilized, I could see 60-something being worth an A. Pravin Vaidya had a mean of 17% and a standard deviation of 14% on one CS 373 (Algorithms 1) midterm at UIUC; a cutoff of 60 of A in that course might have been quite reasonable. I'm guessing it wasn't like that here.

May. 22nd, 2006 06:56 pm (UTC)
Re: Context is everything, and getting there is half the fun
No formal numeracy is required to understand introductory macroeconomics or introductory microeconomics (as taught as KSU, although the 1974 Enyclopedia Brittanica was similarly undemanding). The most demanding skills involved are reading graphs without scales, which is easily superseded by even remedial math classes.

So, math-free. All that is required to pass, is the ability to learn and use the terminology -- the same as facility in picking up foreign languages. I handled those courses as I handled taking French, not how I handled taking math.

Intermediate microeconomics (the other econ course I took) did require some algebraic facility, but nothing beyond college algebra. But it would have been possible to get 85% of the course material without that.

Econometrics (and its spinoffs) are the only part of economics that requires anything beyond college algebra to handle. Mostly multi-variable numerical calculus and statistics.

The verbatim quote from the instructor was "flunk out all business majors". I'm allowing for hyperbole. I think the cut was calibrated to hand out 5% A's. The arithmetic mean was in the low forties. The D/F cut was in the low teens; I half-recall that 20%-30% were getting F's on the exams.

Latest Month

December 2008

KSU Genetic and Evolutionary Computation (GEC) Lab



Science, Technology, Engineering, Math (STEM) Communities

Fresh Pages


Page Summary

Powered by LiveJournal.com
Designed by Naoto Kishi