Schools usually reach for online tests for one of two reasons: printing has become a bottleneck, or they want marks the same afternoon. Both are good reasons. What stops most schools is the assumption that computer-based testing needs a licensed CBT platform, a dedicated lab and an IT vendor.

It does not. What it needs is a browser, one device per student for the duration of the test, and a delivery layer that takes exam security seriously. This guide covers what that means concretely and how to run your first paper-free test without embarrassing yourself.


When online delivery beats paper

Good fitTopic drills, weekly quizzes, revision tests, remote or hybrid classes, and anything where you want marks and analytics within the hour.
Poor fitMore students than devices, unreliable power, or a full-length mock that must mirror a pen-and-paper entrance exam.

The device ratio is the decisive constraint. If two students must share a screen, you are running two sittings or accepting a compromised test — paper OMR is the better tool at that point.

What exam security actually means

“Proctoring” is sold as if it were one feature. In practice it is a stack of independent controls, each closing a specific hole. None is sufficient alone; together they make casual cheating more effort than studying.

ControlWhat it stops
One single-use link per studentLink sharing, sitting a test twice, someone else opening it later
Server-side timerClosing the tab or pulling the network to buy thinking time
Fullscreen enforcementA second window with notes or a chat open beside the test
Tab-switch detection with strikesSearching for answers mid-test; repeated switches auto-submit
Copy, paste and right-click blockedLifting the question text into a search engine or a group chat
Question and option shufflingReading answers off the neighbouring screen
Per-attempt event timelineNothing by itself — but it lets you review what happened afterwards

Two honest caveats. A determined student with a second phone under the desk is not stopped by browser controls — invigilation still matters. And strike thresholds need announcing in advance, because a genuine accidental tab switch should not silently cost someone their paper.

Setting up a test

01

Create the test and set the rules

Title, duration, marks per question and negative marking if you use it. Decide up front whether questions shuffle, whether students may move backwards, and how many violations trigger auto-submit.

02

Add questions

Four objective formats cover almost all classroom testing: single-choice, multiple-answer, true/false and numeric. All four grade deterministically, which is the whole point of using them.

03

Assign students

Each assigned student gets their own link, tied to their name and roll number. Generate them before the exam, not during it — distributing links is the part that runs late.

04

Publish, then hold

Publishing makes the test live. Keep the links until the room is settled; the timer starts when a student begins their attempt, not when you publish.

Test day

  • Do a two-minute dry run with a throwaway test a day earlier. It surfaces the browser, popup-blocker and lab-permission problems while they are still cheap.
  • Explain the lockdown before you start. Students should know that leaving fullscreen is recorded, that repeated switches auto-submit, and that the clock does not stop.
  • Have a paper fallback ready. A printed set of the same questions costs nothing and turns a power cut from a crisis into an inconvenience.
  • Watch the room, not the dashboard. Browser controls handle the software surface; you handle the physical one.
  • Answers save as they are chosen. A dropped connection means reopening the link, not starting again.
Announce the strike rule. Tell students the exact number of violations that auto-submits, and that alt-tabbing to check the time counts. Surprise enforcement is what turns a fair control into a parent complaint.

Results and analytics

Objective questions grade the moment a student submits, so the useful work starts immediately rather than after an evening of checking. What you should expect to see:

  • Scores and percentages for the whole cohort, sortable, exportable as CSV
  • Per-question accuracy, so you can spot the question everyone got wrong
  • A per-attempt proctoring timeline — start, focus losses, fullscreen exits, submit
  • Distribution across the class, not just the average
  • Results in the parent and student view, next to that student’s paper exam marks

That last point is what makes online tests worth adding rather than switching to. In ScanScore the same account scans paper OMR sheets from a phone and runs browser tests, and both land in the same teacher, school and family dashboards — so a term’s assessment history is one list. The web dashboard section shows how those views differ by role.

Readiness checklist

  • One working device per student, charged, with a modern browser
  • Power backup, or a session short enough to survive a cut
  • Network tested at the same time of day, in the same room, with all devices on
  • Links generated and distribution method decided in advance
  • Lockdown rules announced to students and, for younger classes, to parents
  • A printed paper version in the drawer
  • Someone who is not invigilating available to fix a login

Work through that list once and online testing stops being an event. If your device ratio is the thing standing in the way, the paper route is still fast — scanning OMR sheets with a phone gets a class of 60 graded in a few minutes with no devices at all.