An online exam is defined before anyone answers a question — in the settings. Get the twelve below right and the exam runs itself; get them wrong and no amount of invigilation fixes it.
The 12 settings
| # | Setting | The decision |
|---|---|---|
| 1 | Exam type | Objective, subjective (typed answers), or mixed |
| 2 | Question types | MCQ, true/false, fill-in-the-blank, image-based, analytical |
| 3 | Marks per question | Uniform, or variable (e.g. 1–5 marks by question weight) |
| 4 | Duration & timers | Total minutes — optionally a per-question timer (e.g. 30 seconds, then auto-advance) |
| 5 | Result visibility | Instant score + analysis, or withheld until official release |
| 6 | Navigation | Free movement between questions, or forced sequence |
| 7 | Negative marking | Deduction per wrong answer, or none |
| 8 | Syllabus | Subjects and topics the paper draws from |
| 9 | Difficulty mix | Which levels (easy → very difficult) and in what proportion |
| 10 | Identity capture | Webcam photo capture and verification of the remote candidate |
| 11 | Randomization | Unique question set and sequence per candidate |
| 12 | Passing marks | The cutoff that flags eligible candidates automatically |
The settings that do the security work
Three of the twelve carry most of the integrity load:
Layer the rest of the exam security stack on top — secure browser, activity logging — for high-stakes exams.
The settings that do the pedagogy work
Result visibility and difficulty mix decide what the exam teaches. A test series shows instant scores with weak-area analysis, because feedback is the point; an entrance exam hides them, because fairness is. Difficulty mix drawn from a well-tagged question bank is what makes two papers equivalent even when randomized.
Design flow in practice
- Build the question bank with subject, topic and difficulty tags.
- Define the blueprint — the 12 settings above.
- Let the role structure validate the paper before it's assigned.
- Schedule, assign candidates, go live.
Eklavvya, the assessment engine from the ePravesh family, exposes every setting above per exam — used by universities and training institutes for entrance tests, internal assessments and test prep. Book a free demo to design your first exam.
Frequently Asked Questions
What settings define an online exam?
Exam type and question types, marks per question, total duration and per-question timers, navigation mode, negative marking, syllabus (subjects/topics), difficulty mix, randomization, result visibility, passing marks and identity capture.
When should instant results be shown to candidates?
For practice and test series, always — instant analysis is the learning value. For entrance and recruitment exams, disable result view so scores are released only after the official process completes.
What does question randomization do?
Each candidate receives a different selection and sequence of questions drawn from the bank against the same blueprint — papers stay equivalent in difficulty while answer-sharing between candidates becomes useless.
How is candidate identity verified in an online exam?
The webcam captures and verifies the candidate's photo at login and during the exam — remote proctoring confirms the registered candidate is the one appearing, from any location.