What is an online exam system?

An online exam system is software for creating and delivering tests in the browser. Teachers write questions and set a time limit, an availability window and the number of attempts, and students take the test on a computer, tablet or phone. Objective questions such as multiple choice are scored automatically, open answers are graded by the teacher, and students can see their results and feedback online.

How can you prevent cheating in online exams without proctoring software?

Design the test so that looking things up helps little, for example by asking students to apply knowledge to new cases instead of reproducing facts. Combine a realistic time limit, a short availability window and randomized question order. Use varied question types, including short written answers, and announce that some students may be asked to explain an answer orally. Hold high-stakes exams in the classroom if needed.

Which question types can be graded automatically?

Multiple-choice and true/false questions can be scored automatically, because each has one fixed correct answer. Fill-in-the-blank questions often need a teacher to check spelling variants and synonyms, and short answers and essays require manual grading against clear criteria. For most tests, a practical balance is mainly automatically scored questions plus one or two open questions that show how students apply what they learned.

How long should the time limit for an online test be?

Ask a colleague who knows the material to take the test and use one and a half to two times their completion time as a starting point. Allow more time for reading texts, listening tasks and written answers than for single-word answers. The limit should be enough for a well-prepared student but too short to research every question. Agree on extra time for students who need it before the test opens.

How do you create a placement test for a language course?

Cover a wide range of difficulty, from beginner to advanced, so you can see where a student's knowledge stops. Combine grammar and vocabulary questions with a reading or listening task, and map score ranges to levels such as the CEFR levels A1 to C2 instead of setting a pass mark. Add a short speaking conversation after the written part, because speaking is hard to test with written questions.

Online Exam System: How to Create Fair Tests and Quizzes

An online exam system lets you create tests in the browser, open them for a set period, score objective questions automatically and grade open answers with written feedback, so students can take a placement test, a quiz or a final exam from any device. Whether those tests are fair depends less on the technology than on the design: a clear purpose for each test, a sensible mix of question types, realistic time limits and honest measures against cheating.

In short

Decide first what a test is for: placing students, checking progress, confirming that a course goal was reached, or practice. The purpose determines length, settings and how you use the results.
Combine automatically scored questions with a few open questions that the teacher grades, so you test both knowledge and the ability to use it.
Time limits, availability windows, limited attempts and randomized question order make cheating harder; open-book design and oral follow-ups make it less useful.
A test is only finished when students have received their results and feedback they can act on.

Four test types and what each one is for

Most course providers need four kinds of tests. Mixing up their purposes is a common reason why tests feel unfair to students.

Placement tests

A placement test decides which level or group a new student joins. It should cover a wide range of difficulty, from easy to hard, so you can see where a student stops being comfortable. There is usually no pass mark. Instead, you map score ranges to levels, for example in a hypothetical 60-point test 0–20 points for A1, 21–35 for A2 and so on. For language schools, a short speaking conversation after the written part makes placement much more reliable.

Progress tests

Progress tests check what students have learned in the last unit or month. Their main job is diagnostic: they show teachers and students what needs more work. Keep them short, focus on recent content and use the results to plan the next lessons.

Final exams

A final exam checks whether a student has reached the course goal and often decides whether they receive a certificate. It needs a clear pass mark, a fixed window, limited attempts and careful grading. An internal final exam does not replace an official exam such as IELTS, TOEFL or Cambridge English. If your course prepares students for one of those, a mock exam in the same format is more useful than a generic test.

Quizzes

Quizzes are short, low-stakes checks, often a few questions after a lesson or a video. Allow several attempts and show results immediately. Their purpose is practice and recall rather than grading, and they work well inside an online course; see our guide to setting up e-learning.

Choosing the right question types

Each question type measures something different. Multiple choice is quick to answer and to score but mostly tests recognition; an essay shows how a student thinks but takes time to grade. A good test uses several types.

Question type	Best for	Grading	Tip
Multiple choice	Vocabulary, grammar, facts, reading comprehension	Automatic	Base wrong options on typical student mistakes, not on obviously silly fillers
True/false	Quick checks of understanding, warm-up quizzes	Automatic	Use it for low-stakes checks only, because a pure guess is right half of the time
Fill in the blank	Grammar forms, terminology, spelling	Manual: the teacher checks accepted variants	Give enough context that only one answer makes sense
Short answer	Definitions, short explanations, applying a rule	Manual	Write a model answer with the key points before you start grading
Essay	Writing skills, argumentation, case analysis	Manual	Publish the assessment criteria together with the task
Question with audio or video	Listening comprehension, observing a process or situation	Depends on the question type used	Test the media on a phone before the exam opens

Automatically scored vs manually graded questions

Multiple-choice and true/false questions are scored the moment a student submits, which is ideal for quizzes and large groups. Open questions need a teacher, but they show what students can actually do with the material. A practical balance for a progress test or final exam is mostly automatically scored questions plus one or two open questions. Plan the grading time when you schedule the test, and tell students when to expect their final result.

Audio and video in questions

Media make listening and observation tasks possible online. A language school can play a short dialogue and ask multiple-choice questions about it; a technical trainer can show a video of a procedure and ask what went wrong. Record audio at natural speed with clear sound and keep clips short. Assume that students can replay a clip unless your tool limits it, and design the questions with that in mind. Speaking is hard to test with written question types, so a short live conversation remains the better method.

Settings that make a test work

The same questions can make a fair or an unfair test, depending on the settings. Decide them deliberately for each test type.

Time limit

A time limit keeps results comparable and reduces the chance to look up every answer. Ask a colleague to take the test and use one and a half to two times their completion time as a starting point. Reading texts and listening tasks need more time than single-word answers.

Availability window

The start and end date decide when students can begin. For progress tests, a window of several days lets working adults choose a moment that suits them. For a final exam, a short window, ideally one fixed session time for the whole group, reduces the chance that questions are passed on. Make sure the window does not close before a student who starts late can use their full time limit, or tell students the latest start time.

Attempts and retakes

Quizzes can allow several attempts, because repetition helps learning. Placement tests and final exams should normally allow one attempt, with a clear retake rule for students who fail a final, for example one retake after a review session. Tell students before the test how many attempts they have.

Randomized question order

Shuffling the question order makes it harder for students who sit together or chat during the test to compare answers question by question. If some questions belong together, such as a reading text followed by its questions, keep them self-contained or split the sections into separate tests.

Pass mark and showing results

Set a pass mark only where a decision depends on it, typically a final exam linked to a certificate, for example 60 percent of the points. Show quiz results immediately. For final exams, consider releasing results only after the window has closed, so that early finishers cannot pass answers on to others.

Fairness and academic integrity in online tests

In an unsupervised online test, assume that students have access to their notes, search engines, AI chatbots and each other. Remote proctoring and webcam monitoring try to prevent this, but they raise privacy concerns and are not foolproof. For most course providers it is more effective to design tests in which looking things up helps little, and to combine several modest measures.

Open-book design: ask students to apply knowledge to a new case, text or problem instead of reproducing facts that a search engine finds in seconds.
Realistic time limits: enough time for a student who knows the material, too little to research every question.
Randomized order and a short window: make sharing answers question by question impractical.
Varied question types: a short written answer in the student's own words is harder to copy than a letter in a multiple-choice question.
Oral follow-ups: a two-minute conversation in which a student explains one of their answers quickly shows whether the work is their own. Announce in advance that this may happen.
Clear rules: say before the test what is allowed, such as notes, dictionaries or calculators, and what is not.
Proportionate stakes: if a test decides a certificate with real value, consider holding it in the classroom and using the online system for delivery and scoring.

Fairness also means consistency. Every student in a group should get the same conditions, the same criteria and the same grading standard. Grade open answers against written criteria and, where possible, grade one question for all students before moving on to the next.

Accessibility: tests every student can take

An online test should measure the skill you want to measure, not a student's internet connection, device or eyesight.

Devices: check that the test works on a phone, because some students will take it there. Avoid long texts that force constant scrolling between the text and the questions.
A practice quiz: let students try the format once without pressure, so the first real test is not also their first contact with the system.
Clear language: write instructions and questions simply. In a math or IT course, the wording should not be the hard part.
Media: offer a transcript or captions where listening is not the skill being tested, and describe images that carry essential information.
Extra time: agree on individual adjustments, such as extra time for students with a documented need, before the exam window opens, and decide how you will provide them.
Technical problems: publish a simple rule for what happens if a connection drops, including whom to contact and how a new attempt is granted.

Feedback after the exam

Results without feedback are a missed opportunity. Students learn most from a test in the days after it, when they find out what they got wrong and why.

Individual feedback: for open answers, write two or three sentences: what was good, what was missing and what to do next.
Group review: look for questions that many students got wrong. Either the topic needs to be taught again or the question was unclear. In both cases, act on it.
Next steps: link every result to an action. A placement score leads to a group, a weak progress test to extra practice and a passed final exam to a certificate.
Privacy: exam results are personal data under data protection laws such as the GDPR. Show each student only their own results and never post a ranked list in a group chat. This is general information, not legal advice.

Tests also support engagement when students see them as part of learning rather than as a verdict. Our article on student engagement strategies has more ideas.

How Zolinex supports online exams

Zolinex includes online exams and quizzes for every course. You can create placement, progress, final and quiz tests with multiple-choice, true/false, fill-in-the-blank, short-answer and essay questions, and add images, audio or video to a question, for example for listening tasks. For each test you set the time limit, a start and end date, the passing mark, the number of attempts, randomized question order and whether students see their results. Multiple-choice and true/false answers are scored automatically; teachers can grade fill-in-the-blank, short-answer and essay answers manually. Zolinex has no remote proctoring or webcam monitoring, so the design measures above remain your main tools for fair tests. Exams sit alongside video lessons, assignments and QR-verifiable certificates. See all features.

A fair online exam starts with a clear purpose and ends with useful feedback. Get those two right, choose settings that match the test type, and the online exam system takes care of delivery and scoring.