August 6th, 2014

Four Key Questions about Grading


There’s an excellent article on grading in a recent issue of Cell Biology Education-Life Sciences Education. It offers a brief history of grading (it hasn’t been around for all that long), and then looks to the literature for answers to four key questions.

  1. Does grading provide feedback to help students understand and improve their deficiencies? The grade itself is feedback, but generally it is accompanied with faculty comments that justify the grade and offer suggestions for improvement. Most of us know the problem here, “The grade trumps the comment,” as one researcher cited says. Students tend not to read the comments; they look at the grade and get on with life. Not all research supports that conclusion. In some studies, students report that they do read the comments but often struggle to understand the feedback, and they don’t always know how to fix what we identified as a problem. As a result, the same mistakes occur in subsequent assignments. Grading feedback is not as effective as we might hope.
  2. Does grading motivate students to learn? Not really. More often, grading motivates students to focus on grades. If learning is part of the equation, it happens more by accident than design. Pass back an exam and everywhere you hear the question, “Whatcha get?” Nobody is asking, “Whatcha learn?” This analysis of grading and motivation offers an even bleaker conclusion. “Grades can dampen existing intrinsic motivation, give rise to extrinsic motivation, enhance the fear of failure, reduce interest, decrease enjoyment in the class work, increase anxiety, hamper performance on follow-up tasks, stimulate avoidance of challenging tasks, and heighten competitiveness.” (p. 161)
  3. Is grading on a curve the fairest way to grade? The practice of doing so started in the early 20th century when it was discovered that IQ scores were distributed across the population in a normal curve. “Conforming grades to a curve held the promise of addressing some of the problems surrounding grading by making the process more scientific and consistent across classrooms.” (p. 162) However, grading on the curve creates other inequities. If you have a bunch of really bright students in one section, some will end up getting C’s while the same raw scores will be B’s in the section where ability is more widely distributed. But most faculty don’t apply the curve all that rigidly. They adjust it, as needed, for a section or a set of exams which erodes the objectivity and consistency. The other problem with the curve system is that it creates competition in the classroom. When students are competing for points, it’s not in their best interest to collaborate or contribute, which pretty much rules out students learning from and with each other. That works out okay for some students, but it’s not fair for those who do learn well with others.
  4. Do grades provide reliable information about student learning? This is the perennial question about what it is grades really measure and if they measure the same things consistently. The research cited in the paper documents inconsistency in grading by individual faculty members (two different grades for the same piece of work when it’s graded at different times) and across individual graders. Rubrics help, but research still identifies unrelated factors that influence grading (like gender, ethnicity, and knowing who the student is, for example). That kind of inconsistency isn’t a problem with objective exams, such as those with multiple-choice questions, but those exams have students selecting answers, which is significantly different than generating answers. That rounds us back to the question of what kind of learning grades really measure.

We grade students to give them feedback, to motivate their learning, to see how they compare with other students, and to measure their learning—all reasonable purposes. “However, much of the research literature [reviewed in this article] suggests that these goals are often not being achieved with our current grading practices.” (p. 163) Yes, that’s a pretty scathing critique, but it’s well documented and our answers to questions this central need to be accurate.

Reference: Schinske, J. and Tanner, K., (2014). Teaching more by grading less (or differently). Cell Biology Education—Life Sciences Education, 13 (Summer), 159-166.

© Magna Publications. All Rights Reserved.

  • David Johnson

    I reads this and ask how you make this relevant in university classes of 200 plus

    I still use – with half TA grading and half my own grading – short-answer tests but the idea of writing detailed comments on 200 papers seems unrealistic

    • David A

      I review every exam with the class, and my classes are much smaller than 200. I find it more efficient to verbally tell everyone the correct answer than writing it on the test of everyone who got it wrong. I also usually offer choices — answer 8 of 10 — so students will hear the answers for the questions they did not answer, presumably the ones they did not know. I briefly discuss the key ideas that need to be in each answer.

  • Paul Desmarais


    I read your Faculty Focus articles and frequently pass them along to faculty here @ JHU. I would offer this as part of the 'grading on a curve' topic. Grading on a curve defeats the entire purpose of mastery education. If grades reflect levels of mastery and you grade on a curve; either up or down (usually up only) some students who have not mastered the material will be represented as having done so, while the level of mastery for all involved is inflated.

    If students are consistently (year after year, semester after semester) failing to master content in a course you (or anyone) teaches, there are probably only a couple of possible causes. 1. The course is attempting to cover or teach more than is reasonable. 2. The class is being poorly taught. 3. The assessment strategy does not accurately measure what students are learning, or 4 (and the most likely answer) some combination of all three of these root causes.

    Addressing failure to master content by inflating grades eases the immediate problem of having to explain how the majority of students failed a given class, but it does not address the underlying or root causes of these failures. And of course, the students still do not know that they are supposed to know.

  • Claire

    So true. I wonder if you've tried using exam wrappers in place of instructor comments. An exam wrapper is a tool that helps students themselves identify some of the trends, etc. that an instructor might otherwise identify for them in a comment on test performance). I learned about exam wrappers from Susan Ambrose, et al, in their book, How Learning Works, but if you just google "exam wrapper," you can learn a lot more about them (and see examples)! Just a thought for the start of a new year … 🙂

  • Russ Hunt

    I'm wondering how you can possibly escape from the clear — though unstated — conclusion of this article: Grading is counterproductive and destructive. The "accurate" answer to "questions this central" can only be: find ways to get rid of it. Yes, it's difficult to imagine education as we've constructed it over the last couple of generations without it — but let's try.

  • Neil

    What I found interesting about this article is the alternatives to grading suggested. The alternatives ease the grading workload and produce room to implement active learning strategies in the classroom. Often active learning strategies are perceived as having an increased marking workload. By questioning the educational value of grading and offering alternative approaches to marking, authors Schinske and Tanner are offering us a way out of the marking and grading grind enabling us to implement active learning in our courses. Kudos to the authors of this great article.

  • perryshaw

    In light of the significance for success in work and life of such things as attitudes, motivations, and relationships (the "affective" domain, or if you like "social and emotional intelligence") grading becomes even more problematic. In point of fact grades only work well for the lowest level of cognitive learning – repetition of knowledge. Grades for analytic, synthetic, evaluative, and creative work are highly subjective, and grades for non-cognitive learning are largely meaningless. Thank you, Maryellen, for more "grist for the mill"; and great suggestions in the referenced article.

  • Guest

    If we remove grading then how does a student fail a class? For some students will not master the content, nor do the work, and so cannot be allowed to pass on. If you have a class with only P/F grading, how will an instructor defend a student's failure? Sounds great until you're the adjunct or lecturer getting fired for a dispute.

    • Russ Hunt

      If the only, or central, justification of this profoundly destructive practice is that we need to be able to fail people, we are in an extremely indefensible position. (I don't quite understand the point about P/F: with ABCDF or 100-0%, how will an instructor defend a student's failure? What's the difference?

  • Pingback: Four Key Questions about Grading | FAU's eLearning Community of Practice()

  • Russ Hunt

    I just reread the conclusion. "We grade students to give them feedback, to motivate their learning, to see how they compare with other students, and to measure their learning—all reasonable purposes." What's strange about this is that you don't mention the _real_ reason we grade: to certify the students' level of knowledge or competence to others, and to facilitate categorizing them. We can achieve all those goals with other practices — but the reason we don't is that we're required to label students with easy, one-variable categories (GPA, for instance). "“However, much of the research literature [reviewed in this article] suggests that these goals are often not being achieved with our current grading practices.” That's really because "grading practices" aren't suited to those purposes. They're suited to other purposes.

  • Pingback: Ready, set, go! Time for another fantastic year.()

  • Pingback: Four Key Questions about Grading | CTL()

  • Vivian

    We use Akindi as a tool for grading. The immediate data allows us to tailor the next lesson to fit key points that were not well understood. In some cases it allows allows TA's to really target the follow-up lessons. I believe they just released a new version

  • Pingback: Four Key Questions about Grading | CTL()