How does "Attempt problems before seeing the solution" work?

A pre-instruction attempt — known as a productive failure attempt — activates relevant prior knowledge and creates a cognitive search that sensitizes the learner to the features of the correct solution. Even failed attempts make the subsequent explanation dramatically more memorable because the brain has already been oriented toward exactly what it got wrong.

How does "Paraphrase instead of copying" work?

Paraphrasing requires you to decode the original, hold its meaning in working memory, and then re-encode it in new language — a process that creates multiple distinct memory traces. Verbatim copying is largely motor behavior: the words move from the page through the hand with minimal semantic processing. Paraphrasing is generating; copying is transcribing.

How does "Self-explain at every step" work?

Self-explanation generates the inferential links between ideas that a text often leaves implicit. Chi and colleagues found that students who spontaneously self-explained as they read achieved dramatically better conceptual understanding than those who did not — because understanding is constructed by building connections, not received by absorbing facts.

How does "Generate your own examples for every concept" work?

Generating a personal example requires you to apply the abstract concept to your own experience — a form of elaborative interrogation that connects new knowledge to an existing memory network. Personal relevance also strengthens encoding through the self-reference effect: information processed in relation to the self is substantially better remembered than equivalent neutral material.

How does "Write what you already know before reading a topic" work?

Prior knowledge writing activates the existing memory network for the domain, making it easier for the new material to find attachment points during encoding. It also surfaces misconceptions you hold, which are then explicitly corrected rather than simply co-existing alongside new accurate information. The generation of imperfect prior knowledge is itself a form of the pre-attempt mechanism that sensitizes memory.

How does "Alternate generation and reading within a session" work?

Interleaving generation with reading distributes retrieval practice across the session rather than deferring it to the end. Each generation attempt also reactivates earlier material, creating within-session spacing. The contrast between what you generated and what you then read is information-rich: gaps are identified while the material is still fresh enough to be corrected.

Slamecka & Graf

What is the generation effect in learning, and how do you use it to study better?

Generate answers, fill blanks, and solve problems — before you are shown the solution

Short answer

The generation effect is the well-replicated finding that people remember information better when they generate it themselves — even from a simple cue — than when they merely read it. It was first rigorously demonstrated by Slamecka and Graf in 1978 and is one of the most reliable levers in cognitive psychology for improving memory.

If you read this sentence carefully, you will remember it reasonably well. If you had to complete "The generation effect says self-_____ beats reading," you would remember it far better. The act of generating — even a single word — recruits deeper processing than passive reading does. Slamecka and Graf documented this effect in 1978, and it has replicated robustly across word types, tasks, and populations. Below are the core practices that operationalize the effect, each with the mechanism and an honest read on where the evidence is strong.

The practices (7)

Why it works

A blank forces generation: the brain must search for and produce the target word rather than just recognize it. That search process creates a more distinctive, elaborated memory trace than reading does. The difficulty of the search, when it succeeds, is precisely what drives the advantage — easier retrieval produces smaller generation-effect benefits.

How to do it

1While studying, write notes with key terms replaced by blanks or underscores.
2Close the source and try to complete all the blanks from memory.
3Only check the original after a genuine retrieval attempt, even a failed one.
4For failed blanks, repeat the fill-in-the-blank process the next day rather than just rereading.

Evidence

RCT / meta-analysis

Slamecka and Graf (1978) demonstrated that generating a word from a rule (synonym, antonym, rhyme, first letter) produced better recall than reading it — the foundational finding. The effect has replicated widely across paradigms.

Honest caveat: The generation effect is strongest for isolated items; benefits are somewhat smaller and more variable when generating within connected text or complex sentences.

— Slamecka & Graf (1978), the generation effect: delineation of a phenomenon, Journal of Experimental Psychology: Human Learning and Memory

Common mistake: Making blanks for trivial words ("the", "and", "a") rather than for the conceptually critical terms — generation is only valuable when it targets the knowledge you actually need to retrieve.

Go deeper on this practice →

Practice this with IX Coach

Reading about a practice changes nothing on its own. IX Coach turns these into a guided, adaptive routine — discerning where you are in real time and walking the practice with you, session after session.

Start with IX Coach →