Corey, Adam and I will assume that you have a fairly large soap box, or else you might fall off it -- it seems rather slippier to me.
I am in total agreement with Adam. Our school board makes the grading scale, our electronic gradebooks are set up with this scale -- WE HAVE NO CHOICE. I would assume this is the case for most teachers today. As soon as I put in an 80 for the grade, my electronic gradebook will say that the student has a B- average. The students and parents have access to the gradebook for that individual student. If that 80% is quality work, I want it reflected in the gradebook. Ergo, scaling. I cannot put letter grades in our gradebook.
I'm much better at teaching and giving quality tests after all these years, but occasionally I do manage to put a problem on a test that is poorly written or that I did not evidently teach well. Students should not be penalized for these mistakes.
Corey said: We've turned one set of numbers into another set of numbers. Why is that second set any better for purposes of assigning grades? I would much rather look at the quality of work on the test and make grade decisions based on that. If a test is more difficult I'll use a different set of cut scores than I would on an easier test.
This statement says to me that you are "scaling" in your own way.
Corey said: What Mark said should be meaningless. We've turned one set of numbers into another set of numbers. Why is that second set any better for purposes of assigning grades?
Isn't this what we do at the AP reading? A student ends up with only 1 of 5 possible scores. Some of the 5's are stronger than others, but we won't know. One of your students might have had 100%, translated to a 5, while another student had a 72, translated to a 5.
From: "Corey Andreasen" <email@example.com> To: "AP Statistics" <firstname.lastname@example.org> Sent: Thursday, May 31, 2012 12:39:27 PM Subject: RE: [ap-stat] Final Grading
The thing that bothers me about systems like this is that it seems to give legitimacy to the totally arbitrary scale that says "90% is an A." There is no legitimacy to such a scale. I'm not picking on Mark here, but I want to point out that we SHOULD find his statement to be totally meaningless! Again, this is not a dig at Mark. It's a dig at the fact that we all understand exactly what he means!
After his formula, he said, "Hence an 80 becomes about a 90; a 50 becomes about a 70; a 100 stays at 100."
We all know that 90 means "bottom of the A range" and 70 means "bottom of the C range." But why? What is so magical about those cutoffs? They are completely arbitrary. Percent means nothing without "of _____." But we treat 90% as if it has meaning.
What Mark said should be meaningless. We've turned one set of numbers into another set of numbers. Why is that second set any better for purposes of assigning grades? I would much rather look at the quality of work on the test and make grade decisions based on that. If a test is more difficult I'll use a different set of cut scores than I would on an easier test.
Converting things to the 90-80-70-60 scale does two things: 1. It removes the responsibility from us for assessing the quality of the work. 2. It reinforces the idea that the arbitrary grading scale is in some way valid, and is therefor dishonest.
When I assign grades I look for clusters that appear in the distribution. Then I skim a few of the papers in each cluster and decide if they look like A work, B work, etc. Sort of akin to "Complete Response, Developing Response," etc. The better scores in a cluster may get a 'plus' and the worse ones a 'minus.'
I enter the letter grades into my grading program and it assigns a number for calculation purposes, but it is not based on a percent of anything.
Now I'll step off my soapbox.
-- Sally Miller T.C. Williams High School Alexandria, VA