← Blog
Click to upload or drag and drop
PDF, DOCX, PPTX, TXT, JPG, JPEG, PNG, HEIC, ODP, ODT, BMP, or TIFF
up to 20MB
Uploading...
To make an open book exam, write questions that ask students to apply, analyze, and evaluate the material rather than just locate and copy it. List the skills you want to test, build problem-based or scenario questions around them, set point values, state exactly which resources are allowed, and give enough time to think rather than just hunt through pages.
An open book exam sounds easier to write than a closed book one. It is not. If a question can be answered by flipping to a page and copying a definition, the exam tests reading speed, not understanding. The whole point of letting students bring their notes is to push the questions higher: into reasoning, judgment, and application. That takes deliberate design.
This guide walks through how instructors and trainers build open book exams that actually measure thinking, what makes a good open book question, how long to make the paper, and how AI can draft the first version from your own course material.
You make an open book exam by deciding what students should be able to do with the material, then writing questions that require those skills with the book open. Start from your learning objectives, not from the textbook chapters. Work through these steps:
If you want a head start on the drafting, an AI exam creator can turn your lecture notes or readings into a first set of questions you then sharpen into open book form.
A good open book exam question cannot be answered by copying a sentence from the source. It asks students to do something with the information: apply a formula to a new case, weigh two competing options, interpret data, or build an argument. The book helps them remember the tools, but the thinking has to come from them.
Strong open book questions usually share a few traits. They are grounded in a specific scenario rather than a generic prompt. They have a clearly defined task so students are not guessing what you want. And they reward reasoning that is visible in the answer, which also makes them easier to grade fairly. For more on writing prompts that hold up, see our guide to writing good test questions.
Open book exams are usually harder than students expect, not easier. Having notes available removes the pure memorization safety net, so the questions can demand deeper understanding. Students who have not actually learned the material spend the whole exam searching for answers and run out of time, while those who understand it use their notes to confirm details and move quickly.
That difficulty is the point. A well-designed open book exam separates students who can use knowledge from students who can only recognize it. The trade-off is that you have to write tougher questions, because anything answerable by lookup becomes trivial.
An open book exam should give students enough time to think and write, not enough to read the entire textbook. A common mistake is assuming open notes mean students need less time, when the opposite is often true: instructors frequently underestimate how long higher-order questions take. Plan the timing by working a few questions yourself and multiplying.
As a rough rule, fewer, deeper questions beat a long list of shallow ones. If students feel they must consult their notes for every item, the exam is too lookup-heavy. Aim for a length where a prepared student finishes with a little time to review, and tell students up front that searching for every answer is a sign they are underprepared, not a strategy.
You prevent cheating on an open book exam mainly through question design, not surveillance. Questions that require original analysis of a specific scenario are hard to outsource or copy, because there is no single answer sitting in the book or online. Define clearly what open book means for your class: which materials are allowed, whether students may talk to each other, and what counts as outside help.
Make the integrity expectations explicit in the exam directions, and consider asking students to sign a short integrity statement before they begin. Creating two or more versions with different scenarios also reduces copying. An exam generator makes it easy to produce alternate versions from the same source material.
The format of an open book exam varies, but it leans toward questions that reward reasoning: short-answer, essay, case analysis, and applied problem sets rather than simple recall. Multiple choice can work if the options are close and the stem requires real interpretation, but pure definition-matching questions are weak in an open book setting because the answer is right there in the notes.
Many open book exams mix formats: a few applied multiple choice or fill-in items to check that students can locate and use key facts quickly, then longer scenario questions that carry most of the points. The key is that the bulk of the grade rests on what students do with the material, not whether they can find it.
Yes, AI can create the first draft of an open book exam from your own material in minutes. Upload your readings, lecture slides, or notes, and the tool generates questions tied to that content, including an answer key. You then do the part that matters most: pushing the lookup-level questions up into application and analysis so the open book format works as intended.
Used this way, AI handles the slow drafting and you handle the judgment. Upload your source file to an AI test generator or build a graded version with an online test maker, then revise the questions until each one asks students to think rather than search. The same workflow scales to closed book papers too, as in our guide to creating a final exam.
Open book does not mean easy to write. Start from what you want students to be able to do, build questions they cannot answer by copying, give them time to reason, and let AI take the first pass at the draft. That is how an open book exam measures understanding instead of search speed.