Grade distributions: What’s appropriate?

I recently received an inquiry in the Center for Teaching and Learning asking for advice on the “appropriate grade distribution” for a class.  In fact, there is not an ideal grade distribution.

The idea of grades, as opposed to narrative reports, were introduced to schooling early in the industrial revolution as a method of sorting and selecting quality (similar to the grading of meat quality). The truth is, there are so MANY variables in the way faculty calculate grades they are far from comparable.  (This was illustrated in a recent workshop I did twice here for OSU in which I placed a series of marks for faculty to calculate and the audience of 80 teachers produced every grade from an F to an A from the same set of marks!) Grade distributions assume not all students will succeed, and that a certain portion of them will fail.  A mathematically generated bell curve compares student to student, rather than student to academic outcomes, or standards of proficiency.  A focus on grade DISTRIBUTION, as well as “GRADING ON THE CURVE”  are therefore inappropriate when it comes to capturing learning.

Instead, we should be looking at PROFICIENCY LEVELS.   When we are clear about performance and knowledge levels, then we can compare students to  program-level (or course specific) outcomes.  Students need to know the proficiency levels so they can hit them.  When proficiency expectations, or clear criteria for academic success, are clearly and consistently communicated to, and supported by the instructor, it is possible that ALL students could work hard, and actually meet the proficiency expectations.  In that case, all students receive the grade of “A:”  the teacher delivered a well designed course, with assignments that were aligned to the outcomes, and the students worked hard to meet those clear expectations.    The questions we need to ask are: what skills and knowledge are targeted for development in our programs and courses?  What levels of proficiency must our students demonstrate at certain points in the program (or course)?  What kind of assessments will best support students in demonstrating their knowledge and skill development?

The purpose of schooling is to educate all students to the best levels possible…not to sort and select them into a statistical model.