How to predict exam questions (without betting on one topic)

If you want to know how to predict exam questions, the honest answer is that you don't guess them: you work them out. Your professor leaves clues all semester (what they repeat, what they draw on the whiteboard, what's in the syllabus, what they ask year after year) and almost nobody writes them down. Put them together and you can tell, with decent confidence, which topics are coming and how they'll be asked.

What you can't do is bet everything on two topics and skip the rest. Predicting questions is for ordering your studying, not cutting it. The point is to know where to start and how deep to go on each topic, not to drop half the syllabus.

I'll walk through it with an intro Management course, a theory-heavy class full of names (Taylor, Fayol, Weber, Mintzberg) where every lecture slide looks a bit like the last one and it's hard to tell what actually matters.

The clues your professor leaves

No professor writes an exam out of thin air. They build it from what they think is central to the course, and that shows up all semester in a lot of ways. These are the ones that tell you the most:

• The syllabus. Read the learning objectives, not just the list of units. If one says "compare classical approaches to management", you've basically got a question already.
• What they repeat. If a concept shows up in three different lectures (say, Fayol's principles, which come back when you cover organizing and again with leading), it matters.
• What they flag. "You need to know this", "I always ask about this", "we'll come back to this". Lines like these are worth more than any study guide.
• What they add beyond the slides. A diagram they draw on the board or an example from their own experience is their way of explaining the topic, and it's often how they'll ask about it.
• How much time they spend. If one unit took three lectures and another took half of one, the exam will probably be weighted about the same way.
• In-class problems and cases. What you practiced in class is what gets tested. If you worked through cases where you had to identify a company's structure, expect one on the exam.
• Required readings. Chapters marked as required carry more weight than the optional ones.

No single clue is enough. What counts is when several of them point at the same topic.

Past papers beat every other clue

Of all the clues, the most reliable is old exams from the same professor. You're not reading between the lines anymore: you're looking at what they actually asked.

How to read them to predict, which is different from sitting them:

1. Collect every one you can and sort them by date.
2. Write down the topic of each question, not the question itself. "Compare Taylor and Fayol" and "How does scientific management differ from classical administrative theory?" are the same topic.
3. Watch the verbs. If it's always "compare", "explain" or "apply to the case", you know what kind of answer you'll be writing.
4. Look for what's always there. In a course like this, there might always be a case to analyze with one of the theorists and a question comparing approaches.
5. Check what changed last time. If the most recent exam switched format (say, to multiple choice instead of short essays), it'll probably stick.

How to use them for practice later, sitting them closed book on a timer, is in how to use past papers. Here you're only using them as a map.

Build a probability table

Once you've got your clues, make a simple table. One row per topic on the syllabus, one column per clue: in the objectives, repeated in lectures, flagged as important, on past papers, practiced in class. Tick where it applies and add them up.

For intro Management it might look like this:

• Classical approaches (Taylor and Fayol): in the objectives, repeated, on nearly every past paper. Five ticks.
• Mintzberg's structures: practiced with cases and flagged as important. Four ticks.
• Weber's bureaucracy: on the syllabus and on some past papers. Three ticks.
• Systems theory: on the syllabus, covered in one lecture. One tick.

The topics with the most ticks get studied first and deepest. The ones with a single tick still get studied, just in less detail: enough to explain them in a paragraph, not a page. If you're short on days, this order decides what you do first; how to spread it across the week is in how to study for an exam in a week.

Write the questions the way your professor would

Knowing a topic is coming isn't enough. You need to know how it'll be asked. The next step is turning each row of your table into two or three questions, using the verbs the course uses.

For the classical approaches, for example:

• "Compare the contributions of Taylor and Fayol to management."
• "Explain three of Fayol's principles and give a present-day example of each."
• "Analyze which elements of scientific management you can see in a fast food chain's kitchen."

Writing them yourself forces you to look at the topic from the side of the person asking, and that's already studying. But the real payoff comes next: answering them without looking. Testing yourself locks material in far better than rereading it, and a major review of study techniques rated practice testing among the most useful.

If you'd rather start from ready-made questions, how to make practice questions from a PDF covers doing it by hand and with AI. That post is about practicing all of your material; this one is about picking what to practice first.

Review sessions and office hours

The last lecture before the exam, or the review session, is the most direct clue you'll get and the one most people waste. Go, even if you think you don't have questions.

What to listen for:

• What other people ask. If three classmates ask about the same thing, the professor notices that topic was hard, and sometimes tests it for exactly that reason.
• How they answer. If the reply to a question is "I won't ask about that in much detail", you know how deep to go.
• What they say about the format. How many questions, whether there's a case, whether it's open book, how much time you'll have.

And one question you can always ask: "Will the exam look like past years?" The answer tells you how much to trust your table.

What not to do

Predicting questions has real risks, and it's better to know them before they bite:

• Skipping topics. A topic that's never come up can come up this year. Everything on the syllabus deserves at least one read and one attempt at explaining it.
• Trusting hallway rumors. "Someone said it's only unit 3" isn't a clue. If it doesn't come from the professor, the syllabus or real past exams, don't use it to cut anything.
• Memorizing answers. If you memorize a model answer to "compare Taylor and Fayol" and get "what would Mintzberg criticize about Fayol?", you've got nothing. Prepare the reasoning, not a memorized answer.
• Using another professor's exams. If the course changed instructors, old exams tell you a lot less.

If your lectures are already in Filmina, a web app that explains your lecture slides out loud, the mock exam in Practice quizzes you on a timer without showing how you did until the end, and if you've set up a study plan with an exam date, you can have it ask only about the files that are on the exam. That covers the last step: checking that you can actually answer what you predicted.

Frequently asked questions

Can you predict exam questions?

The exact wording, almost never. The topics and the format, yes, with decent confidence. You work them out from the syllabus, what the professor keeps repeating and past exams from the same course.

What topics do professors usually put on exams?

The ones in the course objectives, the ones covered across more than one lecture and the ones you practiced with problems or cases. If they also show up on past papers, they're close to certain.

Should I only study the topics I think will be on the exam?

No. Study those first and in more detail, but go over the rest of the syllabus at least once. A topic that's never come up before can show up on your exam.

Are exam review sessions worth going to?

Yes, they're one of the best clues you'll get. You hear which topics everyone struggled with, how the professor answers and what they say about the format, and you can ask whether the exam will look like past years.