[Smt-talk] criterion-referenced grading

[Kris Shaffer]

Zac,
Thanks for your thoughts and questions.

1) I reserve a grade of 4 for near perfection. 3 means "a few small mistakes, but clear understanding of the topic" (from my syllabus—I don't know if I took this wording from Dan Harrison or not, but the 4-point scale and the general meaning of each value are from him; I'm just applying it to multiple categories). Usually the difference between the two is pretty easy to pin down—for instance, two or more illegal parallels in a phrase-length exercise is definitely not 4-worthy; resolving the cadential 6/4 up instead of down will not get you a 4 for cadences; stems going the wrong direction for software; etc.

2)  The numbers 0–4 are neither percentage values, nor do they correspond to letter grades. For the final grade, a preponderance of 3s is a C, half 3s and half 4s is a B, and a preponderance of 4s is an A. Any 2s or lower bring the student down to an overall grade of D, which is failing at CSU (or, more accurately, prohibit going on to the next level in the sequence and require a retake). I'll note that I purposefully designed final projects that the students should have every possibility to do well on—and know where they stand before they turn it in—and which require an individual meeting with me while working on it (analysis, composition, or transcription projects). This should prohibit surprise end-of-term 2s for students otherwise getting 4s. Things that won't work for (like sight-singing exams) are done earlier in the semester, with at least a couple weeks for a student who underperformed on the final one to come back later and attempt it again.

As for grading, we don't have TAs—it's all me. Thus, it's a non-trivial aspect of the system to consider. :) I find grading slightly faster, and less painful in the new system. I'm not worrying about what a single grade will do to a borderline student (or the psychological impact of the letter). It is information to help them as they prepare for the next assignment (or redo the current assignment). I also don't have to worry about giving a B to a student that did most of the assignment well, but failed to do the very thing the assignment was about, just because it averaged out. I can give them three grades of 4 and a 0 or 1 on the thing they missed. They will get the message, and I can still affirm what they did well. Thus, I end up being less fussy about grades, and I think that actually saves me time.

Kris Shaffer, Ph.D.
Assistant Professor of Music Theory
Charleston Southern University
http://kris.shaffermusic.com
twitter: @krisshaffer

On Mar 13, 2012, at 6:23 PM, Zachary Cairns wrote:

> Kris,
> 
> This is a really interesting idea.  I am especially excited by the
> possibility of isolating a student's trouble areas quickly, both for
> me and for them.  It seems like there might be a good deal of
> overhead, in getting the system set up, but the potential benefits are
> worth it.
> 
> I have two questions about how you've been able to make this work:
> 1) Let's say you give a short assignment like, "Write a chord
> progression which ends with a perfect authentic cadence and uses a vi
> chord properly." That assignment might have a grade for something like
> "logical/legal/stylistic chord choice," cadences, vi chord, S-A-T-B
> part-writing (maybe even broken down further, into avoidance of
> parallel whatevers, appropriate doubling, correct notes in the chords,
> etc, etc.), and notation software, if it is to be completed that way.
> I understand that each category would receive a 0-4...but how do you
> determine what constitutes a 3 vs a 4.  I'm anticipating the two
> friends in my class who will say "I only had one more parallel octave
> than Steve, so why is that a 3 and his is a 4?"
> 
> 2) Assuming that your school requires you to submit end-of-semester
> grades with a letter attached to it, so that students can calculate
> their all-important (or so they think) GPA...how do you convert the
> range of 3's and 4's into actual letter grades at the end of the
> semester?  Is it as simple as 4 = A, 3 = B, 2 = C, etc?
> 
> Then, of course, there's the selfish part of me that wonders if you
> find yourself spending more time grading each individual assignment,
> if you are asking yourself which error affects which topic's grade?
> Or, is the method "bulletproof" enough to let a TA do some paper
> grading for you?
> 
> Anyway, it's a really interesting idea, and I thank you for sharing it!
> Zac
> 
> 
> 
> 
> -- 
> ______________________________
> Zachary Cairns
> Assistant Professor of Music Theory
> University of Missouri - St. Louis
> office: 310 Music Building
> office phone: (314) 516-5999
> cairnsz at umsl.edu
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.societymusictheory.org/pipermail/smt-talk-societymusictheory.org/attachments/20120313/58eefaf2/attachment-0004.htm>