How to Create Self-Testing Questions That Actually Work
Writing a self-testing question seems simple: you take a topic, turn it into a question, and that’s it. But most questions students write only check whether they recognize something familiar, not whether they truly understand it. Knowing how to create effective review questions makes the difference between reviewing on autopilot and actually spotting your knowledge gaps before the exam.
Why most review questions fail
When you write a question right after reading a paragraph, your brain still has the information fresh. The result: questions that get answered through shallow recognition, not real recall. It’s the same trap that makes rereading a topic give you a false sense of mastery. This phenomenon is closely related to the signs that you’re recognizing instead of remembering, something worth identifying before designing your questions.
The difference between recognition questions and comprehension questions
A recognition question asks you to identify a fact among familiar options. A deep comprehension question requires you to rebuild the idea from scratch, connect it to others, and apply it to a new case. The first is answered with photographic memory of your notes; the second can only be answered if you actually understood the concept.
- Recognition: “Which of these is the definition of X?”
- Comprehension: “Why does X happen in this situation and not in this other one?”
- Recognition: “Which formula is used to calculate Y?”
- Comprehension: “What would happen to Y if this variable changed?”
How to write self-testing questions step by step
To create effective review questions, follow a concrete process instead of improvising as you go:
- Close your notes before writing the question, not while looking at them.
- Frame the question thinking about what you’d need to explain to someone who knows nothing about the topic.
- Avoid questions that can be answered with a single loose word.
- Always include a “why” or “how” in at least half of your questions.
- Let a few hours pass before trying to answer them, instead of answering right away.
Use verbs that demand thinking, not just remembering
The verb your question starts with shapes the type of mental effort it triggers. “Define,” “name,” or “list” can often be answered with literal memory. “Explain,” “compare,” “predict,” or “justify” force you to handle the idea with fluency. If you notice almost all your questions start the same way, it’s time to vary the verb and raise the level of demand.
Questions that connect ideas across topics
Deep understanding doesn’t live in an isolated topic, it lives in the connections between topics. Ask yourself things like: “How is this concept similar to the one we saw last week?” or “What would happen if I applied this idea to the previous topic’s problem?” This type of question better simulates what a real exam will ask, where questions are rarely labeled by topic.
The mistake of writing too many easy questions
It’s tempting to fill your self-testing list with questions you know you’ll answer well: you feel productive and confident. But that confidence is misleading if it doesn’t reflect your real level of understanding. Self-testing while studying that only confirms what you already master isn’t helping you improve, it’s just keeping you entertained. Save room for uncomfortable questions, the ones that truly make you doubt.
How to know if a question is good enough
Before saving a question to your review bank, try this test: try to answer it out loud without looking at anything, as if you were explaining it to a classmate. If you go blank within a few seconds, the question is good because it reveals a real gap. If you answer immediately with a loose word, it’s probably a recognition question disguised as a comprehension question.
Turn your questions into a system, not a one-off exercise
Writing good questions once doesn’t help much if they’re then forgotten in a notebook. What makes the difference is reviewing them on a spaced schedule, mixing old and new questions to force real recall. This is where Aevoran’s Flashcards tool comes in handy: you create your own question/answer cards, and the system applies automatic spaced repetition based on your correct and incorrect answers, prioritizing exactly the questions that give you the most trouble.
Adapt your questions to the type of content
Not every subject is self-tested the same way. A historical fact or a vocabulary term calls for different questions than a multi-step reasoning problem. If you’re studying purely memory-based content, short and frequent questions work well; if you’re working with reasoning, you need questions that force you to develop the full process, not just the final result. You can dig deeper into this if your subject leans more toward calculation or applied reasoning.
Review and improve your questions over time
Your first self-testing questions won’t be perfect, and that’s fine. Every time a question turns out too easy or too ambiguous, rewrite it. The goal isn’t to accumulate hundreds of questions, but to have a small bank of questions that truly push you to think every time you answer them.
Frequently asked questions
How many self-testing questions should I create per topic?
There’s no fixed number, but it’s better to have a few demanding questions than many shallow ones. Between five and ten well-built questions per topic are usually enough to cover the key concepts and spot real comprehension gaps.
Is it better to write the questions while studying or afterward?
It’s better to write them right after closing your notes or book, without looking at them. Writing them while reading the content tends to produce recognition questions, because the information is still too present in your short-term memory.
How do I avoid making my questions too easy?
Use verbs like explain, compare, justify, or predict instead of define or name. Also, try answering the question out loud without looking at anything: if you answer in seconds with a loose word, it’s probably too shallow.
Does the same type of question work for every subject?
Not quite. Memory-based content benefits from short, frequent questions, while reasoning-based content needs questions that require developing the entire thought process, not just giving a final result.
What should I do if I can’t answer my own self-testing question?
That’s a good sign, not a failure: it means the question detected a real gap in your understanding. Go back to the original material, study that specific part, and answer the same question a few days later to check whether the gap has closed.