CBT exams in Nigerian schools without a lab
Most Nigerian schools that want to move exams onto screens hit the same wall: a computer lab with forty working machines is expensive, and the ones that exist are often shared, ageing, or reserved for the computer studies syllabus. The assumption that CBT requires a lab is what stops most schools from starting.
It does not. If your students can reach a browser - on a phone, a tablet, a borrowed laptop - they can sit a computer-based test. What actually needs planning is everything around the exam: how you schedule it, how you invigilate it, and what happens when the power or the network goes.
Start with one subject, not the whole school
The schools that succeed with CBT almost always begin with a single subject in a single class - usually a junior class, in a subject with a lot of objective questions. One teacher builds one question bank, one class sits one test, and the school learns what its own constraints really are before anything is at stake.
Running a full terminal examination as your first attempt is how schools end up reverting to paper. A mid-term test costs nothing if it goes badly.
Devices: rotate, don't buy
You rarely need one device per student. Two approaches work well:
- Rotation. Split the class into two or three sittings with different question orders. Twenty devices can examine sixty students in an afternoon.
- Bring your own device. For senior classes, many students already have a phone at home. A test taken at home needs different controls (see invigilation below), but it removes the hardware problem entirely.
Invigilation when you cannot see every screen
Paper exams are invigilated by walking the hall. On screens you need a different set of controls, and the useful ones are mostly about design rather than surveillance:
- Shuffle questions and options per student. Two students side by side see different orders, which makes copying slower than answering.
- Set a real time limit that is enforced on the server, not just counted down in the browser. A countdown in JavaScript can be paused by anyone who opens developer tools.
- Draw from a larger bank. If a forty-question test is drawn from a bank of a hundred, no two papers are the same.
- Accept that a home test is an open-book test and write the questions accordingly - ask for reasoning and application rather than recall a student can look up in ten seconds.
Plan for the power going out - it will
This is the single most important difference between CBT in Nigeria and CBT in the countries most guides are written for. Assume, during any given exam, that some students will lose power or network mid-paper.
What matters is that answers are saved as they are given, not collected at the end. A system that only submits everything on the final click will lose a student's entire paper when their phone dies at question thirty-eight. Before trusting any platform with a real exam, test this directly: start a test, answer a few questions, close the browser without submitting, then sign back in. Your answers should still be there.
Also decide your policy before the exam, not during it: how long is a student allowed to be disconnected, and who authorises a resumption? Writing that down in advance turns an argument into an administrative decision.
Marking is where the time actually comes back
Objective questions mark themselves the moment the paper is submitted, which for most schools is the real return - not the paper saved, but the weekend saved. Two things are worth setting up properly:
- Theory questions still need a human. Any system that claims to mark essays automatically is matching keywords. That is fine as a first pass to order the scripts, but a teacher should look before a score reaches a parent.
- Results should flow into the report card rather than being copied there. Re-typing scores from one screen into another is where mistakes enter, and it undoes the time the automatic marking just saved.
A realistic first term with CBT
- Weeks 1-2: one teacher builds a question bank for one subject. Fifty questions is enough to start.
- Week 3: a ten-minute practice test with no marks attached. The point is to find out what breaks - logins, devices, network.
- Week 4-5: a real mid-term test for that one class.
- End of term: if it held, extend to two or three more subjects. If it did not, you have lost one test rather than a term.
Schools that follow roughly this path tend to be running most objective papers on screens within two or three terms. Schools that try to convert everything at once usually stop after the first bad afternoon.