Grade deflation and inflation

This week marks the start of a new academic term, and for many students, anticipation about the grades they received in the past block. Over the past few weeks, DUB has had a number of articles and opinions about whether students should be able to receive full marks (aka, a ten) for a course. In principle, I completely agree that this should be possible. In practice, I admit to not having done so (yet) for any of my students, although some have gotten close. Maybe this means I am a hard grader or that I try to avoid inflating grades, as so many of my colleagues in other universities feel they need to do given the signal it gives to students looking to compete in a corporate or academic marketplace.

What seems missing from this discussion is the variance in grading practices across disciplines as well as whether there is a lower bound to grades along with the supposed upper bound below a ten. First, it makes sense that in disciplines with well-defined questions and answers, say in computer science or mathematics, a ten should clearly be possible. There is little ambiguity about what the "correct" answer is if the code for a computer program works or if the proof of a theorem is complete. 

In my faculty of the humanities, however, it is rare to have assessments that have completely "right" or "wrong" responses. As I wrote earlier about critical thinking, what is at least as important to having appropriate answers is showing how these answers are argued and what evidence is used to support them. Especially with essays and research papers, there are many ways to approach a question, and I am mindful of my own limitations in understanding the vast literature involved. So should students be with their work. 

The flip side to achieving full marks in a course, which students may feel should be possible, is that rarely do students achieve zero marks, which should also be possible. In effect, there is a lower bound in what grades lecturers are comfortable with giving. If a ten should be possible, then a zero should be as well. How many students are willing to accept that tradeoff?

As far as I know, students rarely receive a grade below four, which is a failing grade, but it still allows them to have a resit to attempt a higher grade. Below a four and this is not possible. But here I will disclose a perhaps dirty secret shared among lecturers: unless there are egregious errors, most students will pass. Maybe in part this is due to students making only enough effort to achieve a six, which is the lowest passing grade, but it may also be because lecturers have an incentive to pass students even if they are not quite there. 

Why is that? Given the considerable amount of time spent at the end of an academic term by lecturers on grading final exams and term papers, all of which need to be completed within a short time so the grades can be registered, if we fail students who would then be eligible for a resit this means a lot more work for us. There would be new exam questions and answers to write, arranging a room for the resit, invigilating the exam, and then grading the new assessment all while teaching courses in the new block. As economists like to say, incentives matter. If being more generous to a student by changing their 5.5 to a 6 may make the student happy (even if it may not reflect actual passing knowledge for a course), it also relieves us of additional duties in an already packed schedule. Have I done this myself? Not consciously? Have I been told that this happens? Yes, although how widespread and conscious it is I am not sure. 

So in our discussion about whether there is a grade ceiling below a ten for students, maybe we should also ask if there is a grade floor above four. If there is, this means our zero to ten scale is effectively one that ranges from six to eight, which disadvantages top-performing students from getting recognition for their work as well as failing students by not ensuring that they are mastering the concepts in their courses. In an ideal world, neither a ceiling nor floor would exist, but that also means enforcing homogeneity in grading practices across disciplines as well as overloading lecturers with extra work in a time of budget cuts and uncertainty. This is not an easy problem to solve and one that is unlikely to be solved anytime soon.