Many words have been written, by Mare as well, about digital proctoring. In short, it means taking over someone's computer and webcam in order to prevent fraud during digital examinations. Quite a few people, including members of the Dutch parliament, have expressed their concerns about user privacy, data storage, GDPR compliance and the dangers of utopian thinking about digital solutions.

All of these concerns are justified. But one may very well think that the delicious omelette of perfect fraud prevention justifies breaking a few privacy-related eggs.

It doesn't.

Digital proctoring cannot deliver on its promises, because it can't even cope with the simplest forms of fraud. Universities shouldn't bother with it, for three reasons:

Smartwatch

1) Taking over somebody's webcam and laptop will not prevent the use of other digital tools available to some students. Should they plug in their smartphones to the digital inspection software? Tablets? Smartwatches? The Xbox Live that projects on the white wall behind your laptop screen? Even if the answer to all these questions is “Yes”, there is no way to check if someone actually has done this. Or simply borrowed their roommate's tablet.

2) The webcam has a limited area of vision. Even a digital noob like me has, by now, figured out how to put my notes off screen when I'm making a video for my students.

There have been some suggestions that students should make a 360 degree scan. Poor, poor inspectors, who will have to watch a multitude of cells all spinning around, like jailers in a dizzying panopticon. If a roommate in a gorilla suit shows up on screen number 16 of 35, I'm pretty certain they won't be able to tell.

Also, why assume that a student, after doing the initial 360 frontside handplant to fakie just sits down and does the exam? Five minutes later, the roommate walks behind the laptop and dances the macarena or helps out with the questions. Doing a circle scan every five minutes to prevent this, comes dangerously close to the whirling dervish dance – nobody can do a serious exam this way.

 So, digital proctoring involves a serious invasion of privacy, while only offering very limited ways to prevent cooperation or using online resources. One idea that's doing the rounds now, is to randomly call a couple of students during the online exam. These then get an oral test instead, to see if they can succeed without help and secret access to information.

 This will not do. Not just because it distracts the students and, for some, suddenly changes the way the subject is tested. Not just because this method will only catch a tiny group of really unlucky fraudsters, because of the time it takes university staff to do oral tests. But mainly because I don't have the slightest idea if student Janneke de Wit is one or two meters tall, has dreadlocks or straight hair, is 20 or 30 years old, or why her voice is so impressively low and masculine. I don't really dare ask, either. Should she show her ID? I can't tell if she just pasted a different photo over it, through a webcam.

Elephant

3) Which brings us to the elephant in the room. The digital proctor and even the professor calling in have no idea who is doing the test. Several faculties have had fraud cases involving stand-in students, pretending to be someone else and doing exams for them.

We all know that there are specialized agencies that will write students’ work, even theses. Whenever this ineradicable evil came to light, swift punishment was handed out. Hardened by these experiences, university staff now emphasizes identification procedures during exams, and individual contact during the writing process.

All of this is thrown overboard with digital examination, and the proctoring doesn't replace it with anything at all. We have no way to tell if the person behind the laptop is the actual student who is supposed to do the exam there, then and in heroic solitude. I have taught over a thousand students a course in Roman Law this semester, and I have, sadly, not had the privilege of meeting all of them. This becomes a problem if all of them are doing the test.

One might object that with smaller courses, things will not get out of hand so quickly, because students are less anonymous there. But when our university switches to a 1.5 meter mode, the smaller courses and studies will also have a hard time getting to know all their students by name and by face. This problem will get worse in the future, not smaller.

But, you say, people – and our own well-behaved students in particular – are not that bad and fraudulent in nature. Good news, everyone! Then we don't need to practice proctoring in the first place.

Me and my staff have performed two online tests, recently. A re-examination of a small course, with 31 people, and a large exam for 360 people. The percentage of people who passed was the same as usual, as was the quality of the answers. I don’t have many reasons to think that students, without any form of proctoring, would cheat by using a common WhatsApp Group. If they have done so, it worked against them: a large part had not read one question properly, and wrote an essay about a question that was not asked.

The same applies to using the internet or other resources. A proper “open book” style examination means that unless you already know the answer or how to look it up quickly, you will run out of time trying to digest new information. 

Of course, digital proctoring does make it harder to use resources or helpers. But not much, and it does nothing at all against the real problem of identity fraud. We, as teaching staff, can think of many clever ways to prevent people looking up the answers online or getting help: essay questions, open book examinations, randomizing multiple choice questions. But none of that works if the exam is actually made by their dad, the smart guy from their frat house, or the A-plus Agency, rather than the students themselves.

Against that, there is only one remedy: offline exams, with an ID-check. There's plenty of empty rooms in the university buildings now, so we should be able to do this even with the 1.5 meter rule. If needs be, we can give different exams at different times.

I hope with all my heart that this would be possible. Then, at last, I will meet a few of my students in person.

Egbert Koops is Professor of Legal History at Leiden University