Is Your Grading Model Fair or a Risk Factor?

The processes teachers use to calculate grades can be a major risk factor for students… and institutionally, we need to stop and take a look at whether we are using what the research suggests as being fair.  At Oregon State University, The Center for Teaching and Learning take this type of issue very seriously.  Last year we conducted an experiment.  We provided a room full of university-level instructors a series of assignment grades (equal to a quarter’s work) and asked the teachers to use their current grading system to calculate a final grade.  Final grades ranged from an A to D.  How is this possible?  Surely this wide variance in how teachers calculate student achievement is of concern to our students…and it needs to become more of a concern to us as a faculty.

University of Washington is one of the few PAC 12 institutions addressing the need for greater consistency in the grading process.  They recently established a consistent grading scale. http://faculty.washington.edu/scstroup/Gradescale.html  A common grading scale is a great beginning (91-100 is an A- to A+), but their work does not end there.  The far more complicated issue is ensuring teachers are using FAIR methods of calculating grades.

Here are a list of best practices, based on the preponderance of  research (and statistics) for your review.  And just to give you some foreshadowing…your own OSU Center for Teaching and Learning is designing a user-friendly tool to assist you in selecting a fair and user-friendly grading system for your classes.  Until then, I hope you will take the time to review and apply these 11 tips for fair grading.    If you assign a zero for missing assignments, be sure to read Tip 11!

Grading Tips

1. A grade communicates a teachers’’ culminating judgment about a student’s individual academic gains towards a clearly communicated exit proficiency level.
2. Assigning a “grade” to a student BEGINS with clear learning outcomes. The course outcomes identify WHAT students are to learn. They do not, however clarify the proficiency level students are expected to reach 5by the end of the class. Proficiency levels must be communicated through examples of student work and/or rubrics.
3. Not all classes qualify for the same grading system. Course outcomes dictate the type of grading system appropriate for that class.
4. The validity of a grade is directly related to the quality of the assessment tool. Tests are to be aligned directly to what was TAUGHT and EMPHASIZED during the course (this is called criterion referenced assessment). Exams are not meant to trick students, but rather to accurately capture how much a student has learned as a result of a course.
5. Never grade on the curve. Grading on the curve, unlike criterion-referenced assessment, is “norm-referenced.” In norm-referenced testing “…grades are assigned on how a student’s performance compared with others in the class: Students performing better than most classmates receive the higher grades,” (pg. 327, Nitko &Brookhart 2011). Norm referencing is appropriate when the “population” (the total number of students assessed) is large enough to conduct statistically significant analysis. No courses at OSU are even close to the necessary population to warrant the use of a norm-referencing approach to calculating grades; in the classroom grading on the curve breeds competition and does not accurately reflect what each student has learned.
6. In order to get a fair assessment of students’ learning, it is recommended students be given a variety of ways of demonstrating their knowledge and skills gained. Exams are one method of demonstration; others include graphic organizers, written assignments; presentations; student-teacher conferences, etc.
7. During the quarter, give frequent assessments of students’ academic progress to stay informed about their academic growth but DO NOT calculate all those “dipstick assessments” into the grade.
8. Only grade work that has been announced to students ahead of time. Provide clear expectations for assignments through rubrics on a four point scale: Exemplary, 4 points; Proficient, 3 points; Developing, 2 points; Beginning, 1 point; failure to complete, 0 points.
9. If students MISS an assignment or hand in their work late, give them an opportunity to make it up by completing some other, additional assignment. Clearly remind students of this every time an assignment is made. “If your work is late or missing, you will still be required to hand in that assignment AND watch the “make up video” posted on Canvas, and write a one page reflection.”
10. Students should never be penalized academically for a pedagogical mistake. If no one in the class receives a high mark on the exam, review the test for patterns in students’ performance. If most students miss a specific question, throw the question out; it may have been misleading or was not adequately covered in class.
11. Always give a graded assignment BEFORE the fourth week so students can determine whether they have the necessary prior knowledge or skill level to successfully complete the class. The fourth week is the last week to withdraw from a course without penalty.
12. If you use a percentage method for calculating a grade, never assign a zero to a missing assignment. Instead just assign 59; an F need not be graduated. “when combined with the common practice of grade averaging, a single zero can be a devastating effect on a students’ percentage grade.  The overall grade is unfairly skewed by that one, a typical low score.  To recover from a single zero in a percentage grade system, a student  must achieve a minimum of nine perfect papers…A single zero can doom a student to failure, regardless of what dedicated effort to level of performance might follow (Guskey, 2004; Reeves, 2004 as cited in Guskey, 2015).