eProctoring: How to guarantee identity in remote assessments

Since last year, training centers have had to adapt to the new reality caused by the pandemic. Face-to-face classes have become distance learning, so students and teachers have spent a long time carrying out their activities from home.

This distance format has been possible thanks to the massive use of technology, incorporating to the daily life of teachers and students classes by videoconference, virtual whiteboards and a heterogeneous set of tools for co-creation and dynamization of online activities. Together with a considerable effort from the entire educational community, this technology has made it possible to successfully save the last two academic years.

However, all this digitization of training poses the challenge of assessment, which has also had to be adapted to this new non-face-to-face format. In this scenario, it is essential to guarantee the identity, the absence of help, and the non-impersonation of the person who is taking the test.

In recent years, and in order to guarantee the legitimacy of the tests, different examination models have been introduced and adapted to online assessments as an alternative to traditional ones:

• Final tests sent by email or through learning management systems  (LMS), with a time limit to answer. These systems make it difficult for the student to cheat by finding or sharing answers, but they cannot guarantee that the person taking the exam is who they say they are.
• Tests based on teacher-student videoconference. This test system can be useful for small group assessments, but it is not always possible to adapt it to larger groups or specific subjects.
• Tests with biometric recognition, where the identity and activity of the assessed student is analyzed using Artificial Intelligence.

Tests performed with facial recognition, in which eProctoring technology provides the necessary mechanisms to verify the identity of students during exams, would be a possible solution to the problem of remote assessment.

eProctoring is the set of techniques that enable the verification of the students’ identity and the monitoring and surveillance of their activities to avoid, or at least prevent, cheating in exams performed remotely.

To carry out this monitoring, a wide range of telematic resources (such as web camera, microphone or secure browsers, among others), and biometric methods (facial recognition, keystroke patterns, etc.) are used.

Depending on how the test is monitored, two main types of eProctoring can be identified:

1) Automated eProctoring

In this case, both identity verification and activity monitoring are done automatically, thanks to the use of different Artificial Intelligence and biometric analysis methods.

The information gathered is analyzed, and a report is generated with no need of human intervention.

These automatic systems enable a more economical surveillance, as they accept a greater volume of simultaneous users. 

The negative part of these automated systems is a higher number of situations identified as “cheating attempts” (false positives), which require further verification by the assessors.

2) Live eProctoring

Live eProctoring consists of several systems where surveillance is done by human supervision with the support of Artificial Intelligence mechanisms. The person watching  can see the student through their webcam, have access to their browser or desktop and, in some cases, listen to the student during the exam. Since the same person must be able to watch several students at once, AI ​​systems create alerts that allow focusing attention on suspicious activities detected.

This system causes fewer false positives (the human factor helps to rule out situations where the student is not cheating), but it involves a higher cost.

In this case, it is important to consider privacy aspects, since watchers have access to the image, documentation, and activity of the students taking the tests.

eProctoring tools allow to continue carrying out assessments in a remote training context, helping educational centers to guarantee the legitimacy of the obtained results.

EProctoring helps to reduce cheating but, as it happens in traditional exams with in-person surveillance, it does not ensure the absolute absence of it. However, it allows students to take assessments from home, with some guarantees for teachers that seemed impossible until now.

In future articles, I will explain which factors must be taken into account in order to choose the most appropriate tool for each educational center’s needs, and how the incorporation of eProctoring in the assessment processes can be implemented.