Mastery grading and academic honesty

As we continue to get ready for Fall semester, one of the concerns that I keep hearing is one that I've heard in almost every discussion of online learning since way before Covid-19: That academic honesty is a lot harder, if not impossible, to guarantee in an online course than it is in a face-to-face course. In fact, I've been in meetings in pre-Covid times where academic leaders have dismissed online learning out of hand, simply because of their own personal belief that students will use the online platform to cheat their way through the class. Now that no dismissal of online learning is possible, we need to come to grips with this concern.

On March 15, the night before we threw the switch at my university to make every course online, I wrote this blog post that included this bit:

This is not the time or place to insist on the highest levels of academic excellence, or even airtight mechanisms for ensuring academic honesty. Yes, it's quite possible that students working at home in an online setting could cheat on assignments in ways they may not in a face-to-face setting. [...] There are two ways to respond: Being OK with this, or setting up a mini-surveillance state. The first option is the simpler of the two and so that's what you should go with. Trust students more, and give them more grace and lenience, than you normally do – even more than you are comfortable with. You might be surprised how they respond.

Now that we are coming up on the Second Iteration of the Big Pivot, I would update this in two ways. First, now that we're no longer in emergency mode (despite what it may feel like), we can and should go back to insisting on high levels of academic excellence. Second, there's a third way to respond to the possibility of online cheating other than "being OK" and "police state": Namely, setting up systems that mitigate cheating while still giving students trust, grace, and lenience.

It may seem impossible to insist on high levels of academic excellence on the one hand and at the same time provide trust, grace, and lenience on the other. But fortunately there's a system that gives us the best of both: It's called mastery grading.

Of course if you've been around this website for any length of time, you know that I have been a proponent of mastery grading, a.k.a. specifications grading, ever since 2014 when I perhaps foolishly decided on impulse to redesign my entire upcoming semester based on using specifications grading. Here's a more comprehensive list of some of my earliest posts on this subject. Over the course of the last six years, I've iterated on my use of mastery grading to the point where I think I have a stable framework for implementing it in just about any class I teach, even online. So, I'm being a little cheeky in the previous paragraph.

And yet, I don't think enough instructors address the problem of academic honesty by looking at their systems. We try to treat the symptoms by anything ranging from grave threats in our syllabi to expensive and Orwellian remote proctoring systems (which sometimes fail under pressure and come with significant privacy concerns). But the real cause of the academic honesty problem isn't that we aren't threatening enough or that we haven't installed the right software. The problem is what's always been the problem: Points, and the grading systems based on them.

Think about it: If you're a student, what do you need to do in order to pass a class? Under traditional points-based grading, the answer is: Accumulate a sufficient number of points. Note that I do not say earn a sufficient number of points. We like to think that those points are a natural metric of hard work and actual learning of concepts and processes. But in fact, the very systems in our syllabi that try to connect learning to points, also make it painfully clear that learning is a means to an end — and that end is the accumulation of points. So it's nice if you can actually earn the points through honest means. But when push comes to shove, what matters — as communicated by our own syllabi through points-based grading — is simply their accumulation, through whatever means necessary.

So we set up a deficiency model of learning in our classes by using points as a proxy, similar to the way that in a capitalist society we have a deficiency model of human value using dollars as a proxy. The way you prove your worth in life is to accumulate. The way you prove your worth in class is also to accumulate. And we wonder what drives people to cheat in either setting? And we wonder why so many students find college to be a dehumanizing experience?

If we are really serious about mitigating academic dishonesty, if we are really serious about caring for students and making their Fall 2020 experience an outstanding one, we'll drop the pretense that this is about F2F versus online, and instead take the simplest and best action possible: Get rid of points-based grading and adopt mastery grading instead.

At least three things will improve immediately:  

  1. The incentive to cheat goes away. Under points-based grading, people cheat because it's high-risk/high-reward. You might get caught and face severe consequences, but you might not get caught and accumulate yourself some serious points. On the other hand, mastery grading is predicated upon, among other things, having a robust revision policy for most or all forms of graded work in a course. If you can revise and resubmit just about any significant piece of work — multiple times, and get helpful feedback each time — until you're happy with your grade, then the value proposition of cheating becomes empty.
  2. Student motivation levels rise. By getting rid of points, the narrative about student grades shifts from game-playing ("What do I need to get on the midterm to have a B in the class?") to concept mastery ("I need to study more on Learning Target DC.2") and this leads to students actually connecting with the material as an end in itself, rather than seeing the material as just a delivery mechanism for points. Making a connection to material improves competence. Being able to demonstrate skill in multiple ways — another key tenet of mastery grading — improves autonomy. Competence and autonomy are 2/3 of the ingredients for authentic intrinsic motivation (as I wrote about here).  And intrinsically motivated students are less likely to cheat than extrinsically motivated ones.
  3. Student stress levels drop. The flexibility that mastery grading provides to students means that they don't need to stress about some of the major stressors of Fall semester. If they feel sick and end up missing a day of class where there's an assessment, no worries — just take the assessment at the next session, or set up an oral exam. If they are maxed out with a job and taking care of their family one week and just can't put a lot of effort into an assignment — no worries, just do your best, submit a complete draft, and you'll get feedback and another chance. There's a great power in knowing that your grade is based on what you eventually show that you know.

I'll be sharing details of my mastery grading setup for Calculus and Discrete Structures soon — it's roughed out but I need to clean it up. I challenge you this fall that as long as we're in a situation where everything else is being shaken to its foundations, why not shake more things up in a good way by upgrading your grading? The benefits of doing so may go way beyond improved academic honesty, although that by itself is enough.