Is It Better to Make Your Own Study Guide or Use AI?

The short answer: making your own study guide helps less than most students assume, and the part that helps is not the part that takes the time. In their 2013 review in Psychological Science in the Public Interest, Dunlosky and colleagues rated ten common study techniques and gave practice testing and distributed practice their only high utility ratings. Summarization, highlighting and rereading all landed in the low utility group. Writing a study guide by hand is mostly summarization. So the honest answer is to generate the summary, then do the testing yourself, because the testing is where the learning lives.

That conclusion annoys people, and it annoyed me when I first read the paper, because the four hours spent building a color coded guide feel productive in a way that answering twelve questions badly does not. Feeling productive and being productive are not the same thing, and the research on this is unusually clear.

Is it better to make your own study guide?

It is better to make your own study guide than to passively read someone else's, but worse than spending the same time answering questions about the material. Writing the guide forces you to judge what matters, which is genuine thinking. The problem is ratio: on a typical chapter, maybe twenty minutes of a three hour session is judgment and the rest is transcription, formatting and hunting for page numbers.

What the research actually says

Two findings matter here, and they pull in opposite directions, which is why the question is worth more than a one line answer.

The first is the generation effect, described by Slamecka and Graf in 1978 in the Journal of Experimental Psychology: Human Learning and Memory. Information you produce yourself is remembered better than the identical information handed to you. That is a real point in favor of writing your own guide, and it is why a guide you generated and never looked at again does nothing for you.

The second is the utility ranking from Dunlosky and colleagues in 2013. Here is how the techniques you are choosing between actually scored.

Read down that table and the strategy writes itself. Every high or moderate utility activity happens after the guide exists. Every low utility activity is the guide's construction and its passive review. If you only have four hours before an exam, the research says spend twenty minutes getting a guide in front of you and the remaining three hours and forty minutes answering questions from it.

What making your own study guide is genuinely good for

There are cases where writing it yourself earns the time, and pretending otherwise would be dishonest.

When your notes are a mess and you genuinely do not know what was covered, the act of sorting them is diagnostic. You find the three lectures you never really understood. When the material is mathematical, writing out a derivation by hand is not summarization at all, it is practice. And when the exam is months away rather than days, the extra time cost barely matters and the generation effect works in your favor.

The common thread is that these are cases where the writing is doing cognitive work, not clerical work. Copying a definition out of a glossary is clerical. Deciding that two things you were taught separately are actually the same idea is not.

What it costs you

The cost is almost always measured in questions unanswered. A student who spends Sunday building a beautiful guide for Monday's exam has usually not tested themselves once. They know the material looks familiar. Familiarity is the trap: rereading produces a strong feeling of knowing that does not survive contact with a blank answer sheet, which is exactly why rereading scores low in the table above.

There is a second cost that shows up in professional settings. If you are studying for a certification with a 400 page official guide, hand condensing it is a week of evenings, and the exam does not care how nice your guide looked.

The hybrid method that keeps the benefit

This is what I would actually do, and it takes the generation effect seriously rather than dismissing it.

1. Generate the structure. Upload the chapter, your notes and the slides to a study guide maker and let it produce the concepts, glossary, section summaries, review questions and common mistakes. This is the transcription work, and it is the part with the worst learning return per minute.
2. Answer the review questions before reading anything. Cover the model answers. Attempt every question cold, in writing. You will do badly. That is the point, and it is the single highest value twenty minutes in the whole process.
3. Rewrite the concepts you failed, in your own words. Not copied, reworded. This is where you deliberately buy back the generation effect, spending it only on the material you have proven you do not know.
4. Go back to the source for anything still shaky. The section summaries tell you exactly which pages to reopen, which is the one job summaries are genuinely good at.
5. Retest two days later. That gap is the distributed practice half of the high utility pair, and skipping it wastes most of what step two bought you.

Steps two through five are all high or moderate utility. Step one is the low utility work, and it is the only step you handed off. That is the whole argument in five lines.

When should you write it yourself anyway?

Write it yourself when the material is mathematical or procedural, when your source notes are too disorganized for anything to read them sensibly, when an instructor requires a submitted guide as coursework, or when you have weeks rather than days and want the extra repetition. Generate it when you are working from a clean document, when volume is the problem, or when the exam is close enough that every hour has to earn its place.

Is using an AI study guide maker cheating?

Not for revision. A generated study guide is a reading aid built from material you were assigned, the same category as a summary a classmate shares or a review packet an instructor hands out. It becomes a problem only when the guide itself is the graded work, or when a course explicitly bars AI assistance. If a guide is being submitted, check the policy. If it is only ever seen by you, there is nothing to declare.

Does this apply to teachers and trainers too?

The tradeoff flips when the guide is for someone else. A teacher writing a review packet is not studying, they are producing a document, so the generation effect is irrelevant and speed is the only variable that matters. Same for onboarding, where a policy document has to become something a new hire will realistically read; if the same guide goes out to every new starter, it belongs in a system that tracks who has completed which training rather than an email attachment that quietly goes stale.

What still applies is the review questions. A guide handed out without questions attached invites exactly the passive rereading the research warns about. Whether you are making one for yourself or forty students, the questions are the part that does the work, which is why turning a finished guide into a quiz you can actually sit is usually the next step.

The one sentence version

Making your own study guide is better than reading a stranger's and worse than testing yourself with either, so let the summarizing be automated, keep the questions for yourself, and spread them over more than one evening. If you would rather write it by hand, the step by step method for building one covers how to do it well, and the case for studying by testing yourself explains why step two of the hybrid method matters more than the other four combined.