Monthly Archives: January 2010

Have you ever thought about the pros and cons of our plus/minus grading system versus the A-B-C-D-F system employed by other universities? The plus/minus system actually has several advantages—see for example the reasons RPI adopted a plus/minus system in 2004.

Despite the advantages, under a plus/minus system I’m finding it challenging to implement something I learned from Ken Bain in 2007.

In his 3-day Best Teachers Summer Institute, Bain leads a unit on assessment. In particular he asks educators, “What does it mean for a student to be an ‘A thinker’ in your discipline? What must students be able to demonstrate or do to live up to that standard? How would it be measured if there were no such thing as numerical grades?” Bain suggests that we share with students our answers to these questions to help them understand what it takes to be a so-called A thinker.

When I did Bain’s exercise, I learned that the definitions for A vs. B vs. C thinkers are not trivial, but doable. But consider this. Our plus/minus system has 11 grade categories whereas the A-B-C-D-F system has 5. It’s been 3 years, and acceptable definitions for each of the 11 brackets still elude me. What is it exactly that a C+ thinker can do that a C thinker cannot?

If you have found good ways to define plus/minus grades in the spirit of delineating what students demonstrate they can do (rather than numerical scores), I’d like to hear from you. What works for you?

—–

Here’s another issue with our grading system that borders on triviality, but, well, I think about these things. Suppose in a jumbo lecture an instructor gives 3 multiple-choice exams, averages the scores, and (voilà) assigns final grades according to the chart in the faculty handbook. In our plus/minus system, if we assume the course scores are uniformly distributed between 50% and 100% (which they are not, but bear with me), then 10% of the students are expected to be a half-percentage point below a cutoff, inviting natural rounding up. That is, 1 in 10 students may be arguing for the next higher grade, a grade they technically didn’t achieve. Maybe the number is not exactly 10%, but it’s close for many reasonable grade distributions.

Certainly, there are all sorts of considerations when resolving the rounding question, but that’s not the point. The issue is the sheer number of “rounding cases” that our grading system invites. I recently taught a course where 21% of the final scores were less than a half point below a cutoff.

Happy 2010 everyone.