Learning how to run a secure online exam requires layered controls that protect assessment integrity without creating unnecessary candidate friction. Begin by documenting the stakes of the sitting, the evidence required if an incident occurs, and the support route available to candidates before choosing security settings.
Step 1 — Define outcomes, question types, and integrity requirements. Build or import an approved question bank, assign clear ownership for each item, and decide which candidate-identification checks are proportionate to the assessment. Provide practice instructions before the live window so technical uncertainty is not mistaken for misconduct.
Step 2 — Configure controlled delivery. Set opening and closing windows, attempt limits, sectional timing, navigation rules, and access requirements. Randomise questions or answer order from sufficiently deep banks where appropriate, while preserving a fair and equivalent paper for every candidate.
Step 3 — Select proctoring and monitoring controls that match the risk. OnlineTestPlus includes optional AI proctoring and secure-delivery rules; administrators should confirm the available identity, device, monitoring, and lockdown options for their chosen configuration rather than enabling every control by default. Publish a clear privacy notice explaining what is monitored and why.
Step 4 — Test the complete candidate journey before launch. Run happy-path, connectivity-loss, accessibility, and policy-edge scenarios. Verify that administrators can distinguish technical failures from integrity incidents and that candidates know how to request support without abandoning an attempt.
Step 5 — Review results and integrity signals before releasing scores. Treat automated flags as prompts for human review, not automatic proof of misconduct. Record the evidence considered, apply the organisation’s appeals process consistently, and retain assessment data only for the documented period.