Summative assessment is a cornerstone in education. Once educators have tirelessly taught their students, a final assessment of ability is often administered, and it’s clear why: we need to know how good people are at things. We assign grades to certain intervals of marks not as a badge of honour (although many students will be proud of their achievements), but as a clear way of indicating an individual’s ability to do the thing we assess them for. The high entry requirements to study medicine, for example, and the high standards of assessment throughout ensure that medical practitioners are competent at their jobs. When lives are trusted into the hands of others, we need assurances that they are safe hands.
Fair and accurate assessment is notoriously difficult to do. It requires high standards of consistency between markers of different bias, ability, and alertness. In assessments involving thousands of students, one can see how this problem could precipitate.
Even if we can guarantee that assessments are fair and accurate, an equally complex issue arises: cheating. Students who successfully cheat not only receive grades they don’t deserve, but potentially take away these grades from other students who do deserve them where grading is done on a curve. Cheating can occur in two ways: corruption or subversion.
A corrupt professor could provide certain students with advance knowledge of an assessment, possibly for a fee or other sinister reasons, thus disrupting the balance of fairness. Provided there is no institutional corruption (this is not guaranteed, but is rare), then the corruption can only occur in isolated pockets and comes at a risk of severe penalty for the corrupt individual.
Students attempting to undermine the examination process by finding ways to cheat is the real concern of the majority of educators. Until now, we’ve had an effective (but not perfect) way to increase the risk and difficulty of cheating. Invigilation of examination venues puts students in a room where it is difficult to conceal and interact with illegal material without getting caught. Even so, invigilators can fatigue and miss things, or be fooled by sleight of hand. There is a cost to this of course, invigilators need to be paid (and trained) and venues need to be hired.
The education sector had been relatively happy with this system. A student would have to be particularly enterprising to successfully cheat in a meaningful way to give themselves an advantage and there’s not an educator I know who hasn’t dealt with cheating in some form. This method certainly catches enough cheaters, the ratio of which we couldn’t possibly know, for the system to have remained in place for decades. Then a global pandemic happened...
Proctoring was hard enough before a global pandemic, but now things have become so much more complicated. Adding technology, though useful in principle, can have many unforeseen consequences. With everything connected to the internet, all of the world's knowledge is at one's fingertips. Access to this knowledge turns examinations into “open book” assessments which obscures traditional assessments.
In addition to this, educators are often worried about students:
- Searching for the answers
- Collaborating with other students to work out the answer
- Having other people in the room that can help
- Having their notes on hand
When done well, digital proctoring can offer many benefits:
- Ability to take exams anywhere in the world
- Use of webcams and additional cameras to ensure a fair environment
- Tracking of IP addresses to ensure exams are being taken from the right locations
- Monitoring of screen, keyboard, and mouse activity to ensure the student isn’t accessing information that they should not
- Use of face recognition to check ID’s match the assessment taker
- Use of machine learning to monitor behaviours and actions of the student and flag up suspicious activity
Unfortunately, it's not always that simple. Software solutions are frequently far from perfect, and in some cases may even be malicious. Recently, there have been cases of software having security leaks. Additionally, when software has such extensive access to the user's device it is possible to gain access to data that students wouldn’t consent to giving. This becomes rather a murky issue when examining institutions require the use of certain software for their exams and, as there is no real alternative, students are forced to comply. Not to mention the social issues of facial recognition which could be misused by the software providers or hackers.
Many examinations are structured in a way which enables assessors to differentiate students by ability. A large portion of this consists of glorified fact recollection, commonly known as “bookwork”, which doesn’t require students to exercise their ability to think about their subject. This allows examiners to differentiate between weaker students’ ability to retain standalone knowledge, but not stronger students. There is an argument to be made that this style of assessment doesn’t serve much purpose in the wider context of why we seek to educate people. We should ask ourselves very carefully if we really care about fact retention, especially if it doesn’t give us a true understanding of an individual’s ability. Given our reduced capacity to proctor for fact look-up, perhaps it’s time to do away with it.
An alternative to this is to change, rather than adapt, the way we assess students. By asking questions that force students to articulate and give longer, more nuanced responses, we can use pre-existing anti-plagiarism software effectively. It is even possible to detect changes in writing style, so if someone else was writing answers, then it can be detected with reference to the entire portfolio of work a student has previously submitted.
Opposing fact recall as the differentiator of the lower profiles of marks would then open up our examinations to truly challenge students and provoke higher thought. Reducing the amount of content in subjects would then open up more time to focus on actually understanding it, allowing examinations to probe higher level brain function in more detail. Exam boards can, of course, still scale marks to award the “intended” fraction of top, middle, and lower grades.
No matter how we approach this, proctoring isn’t going away any time soon. It’s important that whatever approach is chosen, students are our foremost thought and not considered an inconvenience. If the ultimate goal of education is to reliably produce competent graduates, then we should always be looking for novel ways to enhance assessment practice in an evolving society.