You are viewing supergee

 
 
15 January 2014 @ 06:05 am
The bad part of teaching  
deGrading

Thanx to Chronicle
Tags:
 
 
( 12 comments — Leave a comment )
Marissa Lingenmrissa on January 15th, 2014 12:20 pm (UTC)
I totally understand the problem, but his solution is terrible. Class rank has no more to do with whether someone has actually got a solid grasp on the material than the most inflated of grades--less, even.

The extreme example, of course, is my Stat Mech class when I was a senior. There were two people in it, me and Pat, and we both got A's, not because of grade inflation (that professor rarely gave A's), but because we both knuckled down and learned Stat Mech. Under the class ranking system, one of us would have to be first and one last in the class, even though we both learned Stat Mech at a proficient level.

Or take Advanced Lab. There were ten of us in that class, and I estimate that eight or nine of us learned modern lab techniques to a proficient level. Should half of us have given up because there were other people in the class who were better at lab techniques? Or in a worse class, should the top half coast and not learn most of the material, knowing that there were plenty of people who would learn even less?

Further, class ranking would still require some form of grading. So it's not even a solution that addresses the problem. It may be that some professors would simply adore to read a stack of student papers and provide them with feedback if only no letter grades were involved, but given what I hear from my professor friends during their major grading periods, I strongly doubt it.
fjmfjm on January 15th, 2014 01:01 pm (UTC)
Indeed.

Also, all of these cries of Doom are bad maths.

Oxbridge used to take X% of Y% of school students selected almost entirely from two types of schools, grammars and Public schools where students stayed until 18.

X has multiplied by about a factor of 10. Y has multiplied by a factor of at least 200.

So Oxbridge is now actually in a position to select far better students.My partner for example was accepted in 1965 with low grades, and would have been rejected by the 1990s.

In that context unless the grades are comparative within the group, I'd expect their students to cluster in As.

At my university, which also has a large pool and selects within a specific band, grades tend to cluster around B/C with outliers in either direction. There is no smooth curve.


Marissa Lingenmrissa on January 15th, 2014 01:12 pm (UTC)
There are so many problems that result from people not grasping the essential fact that, gosh, we have a lot of humans now, no, really, a lot. On the surface it seems like this should be trivial, but really it's not. Whatever your distribution is, normal or otherwise, dumping ever more people into it changes how things are handled, and how they should be.
fjmfjm on January 15th, 2014 02:05 pm (UTC)
It does occur to me tho: one reason why US students may complain more about their grades than UK students may precisely be because they come from a less selective school system. So even if the grading at school is completely fair and on a standard of absolutes they are far more likely to have been near the top of their class. Uni where there are lots of people smarter than them is a shock.

In contrast (just for annecdata), I, a pretty damn sharp kid, was never top.
Marissa Lingenmrissa on January 15th, 2014 02:19 pm (UTC)
Yes. Most US students come from a school system that is selected by their parents' postal code, which is not particularly selective for talent at all. It's not only university/college that's a shock in that situation: kids get used to being the top in their class in grade school and found their identity on it, and then go to a bigger junior high/middle school and have the shock at that stage; iterate for high school.

I think this encourages kids to form identities based on ranking in ways that are unsustainable, and then to give up on themselves as one of the smart kids because they're no longer the smart kid. It also encourages resentment among those who have the most in common. One of the girls who became a friend of mine in late junior high/early high school told me later that she absolutely hated me when we met when we were 12, because she had been the Designated Smart Kid and I had taken her place. On the other hand, it makes for some lonely bits: I only really believed in other kids loving books like I loved books because writers had written characters like that.
blue shark of friendlinessckd on January 15th, 2014 03:32 pm (UTC)
This is stack ranking by another name, and a quick search for that term in conjunction with "Microsoft" will tell you more than you need to know about why it's a terrible idea. (Sure, college students who don't work well together won't hurt the school in the way that employees who won't work together harm a company; that doesn't mean they'll learn well under the circumstances.)
Marissa Lingenmrissa on January 15th, 2014 04:27 pm (UTC)
I actually do think that college students who don't work well together will hurt the school. I think that they'll share in a far less satisfactory learning environment.
Kalimac: puzzlekalimac on January 15th, 2014 03:53 pm (UTC)
Surely reporting class rank is not the same thing as grading on a curve. The problems you mention pertain to the curve. The problem with curve grading is that the individual's grade can then be taken out of the context of the class's size and proficiency. Class rank, especially as it's not judging the individual students, not so much.
Marissa Lingenmrissa on January 15th, 2014 04:27 pm (UTC)
See ckd's comment above. "Student X was the 12th most proficient out of 15 students" tells no one anything about whether they were all quite proficient or none of them were. I don't see where class ranking gives anyone any idea of overall proficiency.
Kalimackalimac on January 15th, 2014 04:47 pm (UTC)
It's inherently relative. A "C" is inherently absolute. This is even clearer when using your example of a 2-person class. You said that "Whatever your distribution is, normal or otherwise, dumping ever more people into it changes how things are handled." Class ranking tells you that. It's only when you draw lines across it and assign grades that you get stack rankings, i.e. the curve.
nancylebov: green leavesnancylebov on January 15th, 2014 04:49 pm (UTC)
" my colleague told the student that in this course he was simply a C-plus student."

This is disgraceful-- the kind of statement has a noticeable risk of causing students to give up, even though it does cause others to try to prove the teacher is wrong.

It's much better to say something like "You've been turning in C+ work".
msrat1900 on January 16th, 2014 05:46 pm (UTC)
The best teacher I ever had gave every grading event (quiz or test) twice. He said "I will not record the results of the first, but you have a week (or two weeks) to see where your knowledge is deficient and to remedy the situation. I will give only two grades on the retest, Excellent and Fail." In fact, the grades we received were either As or Bs, depending on whether we knew it perfectly, or knew it perfectly and were able to discuss it cogently. We all learned the material and, equally important, we learned how to make testing a useful tool for learning.

Education as a whole would be greatly improved by adopting this practice.
( 12 comments — Leave a comment )