Examopotamia1. Decide what the test is for
Write one sentence: “People who pass this test can …”. For example, “can build a pivot table and use VLOOKUP”. Every question should help you decide whether that sentence is true for the person taking it. Anything that doesn't belongs in a different test.
2. List the topics and how many questions each gets
Split the goal into three to six topics and give each a share of the questions that matches how important it is. This list (a “test blueprint”) stops a test from accidentally being 80% about the easiest topic.
3. Write clear questions
Use one idea per question, plain language and plausible wrong options. Our guide on writing good multiple-choice questions covers this in detail. If you are short on time, start from a free template or let the AI question generator write a draft, then edit it.
4. Choose the length and time limit
For screening and quick checks, 10–20 questions in 10–20 minutes is usually enough. Allow about one minute per choice question and more for questions that need calculation or reading. A limit that is too tight measures speed, not knowledge.
5. Set a pass mark
Start around 60–70% and adjust after the first group. If almost everyone passes, the test is too easy or the bar too low. If good people fail, check the hardest questions for unclear wording. Our test score calculator helps you see what a pass mark means in points.
6. Make cheating harder
- Shuffle questions and answer options for every person.
- Use a platform that keeps answer keys on the server.
- Enforce the time limit on the server, not just with a countdown in the browser.
- Draw questions from a bigger bank so not everyone sees the same set.
7. Share it and review the results
Send the link, then look beyond the scores. Question statistics tell you which questions were too easy, too hard or confusing. Fix or remove those and your next test will be better.
Examopotamia handles steps 3 to 7 for you and is free to start.