Robo-invigilation. Does it work? Is it ethical?

Man surveys football pitch through an array of monitors

Covid-19 has accelerated a move to online assessment. Where there is a will to hold those assessments as unseen examinations under controlled conditions, online invigilation – also known as remote proctoring – becomes a possibility.

What is remote invigilation (aka remote proctoring)?

In 2020 SURF (Netherlands organisation to support technologies in education and research) updated its 2016 white paper on remote proctoring. The updates take into account the intensified interest in digital invigilation due to Covid19, the potential and limitations of remote proctoring, and concerns about privacy.

  • SURF white paper on remote proctoring, 2020. Read the summary, and to get a sense of what online proctoring is attempting to eliminate and how it works, read 5.2 of the updated report. For an approach to categorising assessment security, read 5.4.1.

One of the fastest growing invigilation companies is Proctorio. Its critics say that its business model (unlimited use for a flat rate) has led to scope-creep i.e. invigilation increasingly happening with low-stakes and low-risk assessment (see the SURF model above) as well as for the originally-intended high stakes exams which could, for example, gain you a licence to practise as a doctor, engineer or lawyer. Proctorio’s CEO says that the software is now used weekly on 2-4 million students. Actually, he says it’s used “by” them – but there’s the rub.

In September, Proctorio launched a lawsuit against a learning technologist at the University of British Columbia called Ian Linkletter (@Linkletter) for criticising material they had published. Linkletter’s affidavit includes a number of exhibits about inequalities in remote proctoring which are useful in judging software.

What if the algorithms weren’t biased? Should we factor in reductions in privacy and well-being as the necessary side effects of using examinations? After all, stakes have never been higher for students. Higher education is expensive and the jobs situation is currently dire. For these reasons students are often at the front of calls for better assessment security – if there are opportunities to cheat then the degree becomes less respectable and less competitive. Everybody recognises that processes that reward cheating aren’t fair or fit, no matter how creative the cheating.

An alternative question: can the need for controlled conditions be designed out while maintaining assessment security for high-stakes, high-risk assessment, maintaining the breadth of learning associated with unseen exams while also promoting the depth and longevity of learning which typically isn’t?

A choice of readings:

Image source: BBC Sport.

Leave a Reply

Your email address will not be published. Required fields are marked *