The Generation Effect: Why Writing Your Own Quiz Questions Beats Rereading

Most students still reread their notes before an exam. It's comfortable. It feels like studying. But the evidence says you're wasting your time, and that a stranger technique, where you generate the material yourself, builds memory far more effectively.

It's called the generation effect, and it's one of the most robust findings in cognitive psychology.

What the Research Actually Shows

The generation effect has been documented for decades. Slamecka and Graf (1978) ran five experiments comparing memory for words subjects generated themselves with memory for words they simply read. Generated words were remembered far better across all five experiments. The finding has been replicated many times since.

More recent work extends the effect beyond single-word recall. When students create their own quiz questions about a text and then answer them, they recall more on later tests than students who simply reread the material. The questions don't have to be the same ones on the final test. The generation process itself is what makes the material stick, not the specific questions produced.

The key is that generating material forces your brain to do something rereading never does: retrieve, organize, and restructure what you already know.

Why It Works

When you reread, your eyes move over words you've seen before. Your brain recognizes them. Recognition is easy. Retrieval is hard. And it's the struggle of retrieval that makes the memory stronger.

When you write your own question, you have to decide what matters. You have to phrase it so it makes sense. You have to hold the concept in mind long enough to express it in a new way. Each of these steps adds a layer of processing, and each layer makes the memory more durable.

This connects to a broader principle in memory research: effortful processing beats shallow processing every time, provided the effort is directed at the right thing. You're not suffering to no benefit. The difficulty is the point.

There's also the matter of inference. When you generate a question, you inevitably fill gaps using what you already understand. Those inferences create additional memory traces. The next time you encounter the material, you have more hooks to pull it from.

The Rereading Trap

Rereading feels productive. The fluency of seeing familiar words creates a subjective sense of mastery. Psychologists call this fluency illusion. You're mistaking the ease of reading for the ease of recalling.

You can read a chapter three times and feel confident. Put the book away and try to explain a core concept from memory, and the gaps reveal themselves fast.

This doesn't mean rereading is useless. A first read gives you the raw material. But at some point, you need to move past recognition and into retrieval. Generating questions is one of the cleanest ways to do that.

How to Use This Without Overcomplicating It

You don't need to overhaul your study routine. Just add one step after reading: close the notes.

Write a question about what you just read. Not a factual question if the material is conceptual. Better: ask something that requires synthesis. "How does this connect to what I read yesterday?" "What's an example from my own life that illustrates this?" "If I had to explain this to a freshman, what would I say first?"

Then answer it. Write the answer out, even if it's short.

The question doesn't have to be clever. It doesn't have to look anything like a real exam question. The act of generating it, and then retrieving the answer, is what matters.

For subjects with formulas or procedures, try generating your own worked example. Close the textbook, work through a problem from memory, then check your result. The errors you catch in that process are more educational than watching someone else solve it correctly.

A Different Kind of Study Session

This approach changes what a study session feels like. Rereading is comfortable. Generating questions is active, a little awkward, and occasionally frustrating. That discomfort is useful information. It tells you which parts of the material are still loose.

Some students avoid it because it feels slower. You're not covering as many pages in an hour. But coverage isn't the same as learning. A chapter read three times with no retrieval is still just familiar text, not knowledge you can access under pressure.

The generation effect doesn't require any special tools or apps. A blank document works fine. So does a piece of paper. The mechanism is in the generation itself, not in how you capture it.

The Honest Limitation

There is a place for rereading. When you're encountering material for the very first time, reading gives you the scaffolding. You can't generate questions about something you don't yet understand.

The generation effect works best once you have something to work with. Think of it as the second pass, not the first.

Also, generated questions vary in quality. A shallow question ("What is cognitive load?") generates less benefit than a deep one ("Why does reducing cognitive load help learning, and what happens when too much is removed?"). Worth being deliberate about the level of thinking you're asking for.

The Core Idea

The generation effect is simple: you remember things better when you create them yourself rather than receive them passively. It explains why making your own flashcards, writing your own summaries, and formulating your own questions tends to outperform reviewing other people's materials, however well-made those materials are.

Your brain holds onto what it works hardest to produce.

What's the next thing you're going to read where you'll write your own questions instead of just highlighting?