The following entry was contributed by Joseph Michalka, Graduate Associate, Kaneb Center for Teaching and Learning.
You have just finished grading the first test and while there were a couple of high B’s, you were shocked that the average was a 52. With your syllabus saying that the first exam would be a third of the final grade you realize that if you enter the grades as is, then nearly no one will be able to get an A in the course and you are faced with a difficult decision, to curve or not tocurve.
As a teacher, you do not want to be stuck in a situation like this and while there are many ways to prevent such a scenario (see Using Grading Rubrics, Grading Student Work, Fundamentals of Course Design III: Assessment and Exam Design), dealing with it after the fact requires having a flexible grading policy. Grading policies can vary drastically, but most will be contained within the categories of criterion based or reference based assessments. The rest of this post will examine some of the specifics of each and situations where they might be appropriate.
Criterion Assessment
Generally, the criterion method focuses on measuring a student’s progress against stated learning goals or outcomes. As labeled, this grading strategy typically depends on rubrics or criteria that establishes how assignments, projects, and exams will be graded. Depending on the implementation, this can make the grading process extremely transparent which can help prevent questions and complaints. The lack of direct competition has also been touted as helping to increase student cooperation, since the students will not be fighting for a limited number of high grades. Finally, there is no guarantee of a bell distribution of grades as all of the students could end up failing or acing the class, depending on how their developed skills match up against the set learning goals of the class.
Reference Assessment
The reference or comparative assessment is often consistent with the idea of grading on the curve. This method ranks students against each other, typically by normalizing the grade distribution so that the assigned grades match a bell curve with most grades being around a C, or ‘average’ grade. This can be a very effective strategy when students need to be ranked or compared against each other, whether that be for a fellowship opportunity or entrance into a selective program of study. Additionally, this method has been touted as a possible solution to grade inflation (see Princeton’s report on a decade of institution-wide curving). However, a strict following of this method could result in high grades for students who still have not mastered the material, if they happened to know slightly more than their peers or in the inverse situation, students who only missed a couple of points on an exam could receive B’s or C’s if a larger number of students earned perfects. Work by Kulick and Wright convincingly argues that a strict policy of grading on the curve can actually result in false assessments of a student’s capabilities.
Combination Assessments
A real class is complicated and a combination of both assessment methods may be appropriate for your specific situation. As an example, perhaps your department does not require a grade distribution and as such you are fine assigning everyone an A, assuming they earn it. However, you may still prefer to have an internal ranking for your own personal records to help when writing recommendation letters. Another common situation that might require a combination of policies would involve teaching an extremely difficult course, i.e. organic chemistry. The students might only score mediocrely against the course’s stated criterion, but when the difficulty of the class is compared against the rest of the university, a curve might be appropriate so as to not punish them for taking a difficult course.
These brief discussions on grading will hopefully help you start formulating your own grading schema which will likely include aspects from both. Whatever you choose, keep in mind what the assessment process is supposed to provide and aim to accomplish those goals, i.e. feedback, motivation, and a report of the student’s learning. A similar topic focusing on the philosophy of assessment and the perennial problem of grade inflation will be examined in a future blog post.
For more thoughts on grading policies
- http://teaching.monster.com/benefits/articles/5658-why-grading-on-the-curve-hurts
- http://edtheory.blogspot.com/2013/05/curving-exam-scores-is-it-more-harm.html
- http://academics.hamilton.edu/biology/smiller/curve.html
- http://www.gradeinflation.com/
- Kulick, George and Wright, Ronald (2008) “The Impact of Grading on the Curve: A Simulation Analysis,” International Journal for the Scholarship of Teaching and Learning: Vol. 2: No. 2, Article 5.
