Increased flexibility = take at time of choosing (one survey respondent said she can do the exam at home after the kids are in bed – no sitter costs!), less time overall because no transportation time, don’t have to spend time choosing an invigilator etc
Initial testing of the technology was done in the Office of the Registrar to test reliability and the ability of the service to maintain exam integrity. A team simulated a variety of cheating scenarios (someone sneaking in to the room on the floor to hand the exam taker a piece of paper, for example). The team was satisfied that this was a technology that was worth further consideration.
Working group included academics across all faculties, representatives from exam services, office of the registrar and other administrative units, Students with Disabilities, Privacy coordinator, undergraduate student representative.
The other option that we considered was Software Secure. This was not live invigilation. They would review recordings of the exam after the fact.
After selecting Proctor U, the Registrar attended Faculty Council meetings to explain the service and answer any questions. The evaluation criteria includes surveys from Proctor U and our own survey.
We planned a gradual implementation by limiting initial communication to students and slowly increasing the reach of our communications as we became more comfortable that the service was working as intended. This did give us some time to address a few small issues; one regarding information provided to Proctor U that needed to be reformatted as it was initially causing confusion about exam requirements.
Loss of internet service. A student could cause their internet connection to be dropped, review source notes, and then use this information to assist in the exam once the connection was re-established.
While possible, probably not very likely. If it is a concern we could terminate the exam after a certain period of time and force a re-write.
Monitoring the desktop space. Their proctoring method relies on watching the "forehead to chin" of the test taker to look for suspicious eye movements outside of the exam space. What about a student who is doing calculations/making an outline/making notes on scrap paper? How do we make sure crib notes aren't slipped onto the desktop?
They are good at detecting these issues and can do a "scan" of the environment as necessary by asking the student to stop writing and to use a reflective surface to highlight the exam space. It is possible to request that writers have a second camera pointed at their work space. This would provide better security but would be another technical and financial hurdle for the student.
Bathroom breaks. What prevents a student from cheating during a bathroom break?
Nothing realistic can be done to prevent this type of cheating but it is no different from the current situation at remote sites. Through Proctor U, we get exact times of absence from the exam and could note if "sudden inspiration" happened when they returned.
ProctorU proctor can view the test-taker's screen, temporarily borrow the test-taker's mouse and keyboard functions and also maintains a chat window between the test-taker and proctor throughout the entire exam session in the event that the test-taker is in need of assistance.
In the period discussed above, we had 2 incident reports: logout and an unpermitted break
In addition to Proctor U’s survey, we sent our own surveys to students but they tended to complete the Proctor U survey and not ours (our response rate 19.2%)
Now 726 exams to date; we are expecting the quarterly update to reflect similar satisfaction.
Student Union reports as well extremely positive reports from students.
We are currently investigating two issues (only known issues) Reports that invigilators are not always giving additional time for issues that arise Technical issue that our online exam system interprets the chat window with the invigilator as an exam breach and locks student out
Pilot evaluation has been delayed by slower uptake than we expected. We had intended to do evaluation by end of summer but will likely wait until the end of the calendar year. Given the promising results, we will continue to offer the pilot while we complete our evaluation.
Remote Invigilation at Athabasca University
Focused on the future of
REMOTE INVIGILATION AT
Alain May, AVP Student and Academic Services
Richard MacLeod, Registrar
Cindy Ives, Vice President Academic
October 4, 2017
– 3000 exams per month
– paper-based (54%)
– Online through MuchLearning
– Online through IBM Notes
• Current Exam Invigilation
– Free exam invigilation at our
learning centres: Edmonton,
– Invigilation network
– Students can request other
Why Remote Invigilation?
• Reduce costs for students
– A typical external invigilator can cost:
• Up to $CAD 200
• Parking and transportation costs
• Increase flexibility for students
• Reduce exam stress
• Reduce operational time spent on managing
invigilator network, exam bookings and
academic integrity issues
• Reduce physical presence in other regions
• Initial testing of the technology
• Working group
• Selection process
• Consultation and engagement with faculties
• Pilot process with planned evaluation criteria
• Gradual communication of pilot to students
• Loss of internet service – intentional or
• Monitoring the desktop space
– Using unapproved materials
• Bathroom Breaks
How Does It Work?
• Computer with standard webcam, microphone and internet
• Students schedule exam window as per normal with AU and
then book with ProctorU
• During the proctoring process,
– Proctor connects to test-taker via chat, video and audio
connections and can take control of student device
– Proctor maintains chat window with student throughout
– Proctor can stop the exam if misconduct is expected and
request a scan of work area
• Incident reports of any issues (and bathroom breaks) are
provided to us
• Detailed record of all events in session are kept. End to end
recording of the entire session is available for review up to 30
days past write date.
Early Pilot Results
• May – June 15, 101 exams (Survey response 72% in May,
76% in June)
• Mainly Canadian test takers but also international
• Company Service Average – 4.02 May, 4.39 June
• Were you satisfied with the company:
– 46% Strongly Agreed
– 38% Agreed
– 4% Neutral
– 11% Disagree
– 1% Agree
• Proctor Service Average - 4.4 May, 4.57 June
• Please rate your proctor:
– 60% Excellent
– 28% Good
– 11% Mediocre
– 1% Poor
• Evaluation of pilot
– Survey results
– Faculty discussions
– Consideration of
issues raised to date
• Communication of pilot
results and decision