Spaced repetition

Spaced repetition is a study method that spreads review of the same material across increasing gaps of time, combined with active recall, so that you remember more with less total study time. It answers a common problem: people forget most of what they read or cram, then feel they must start over. Used well, short daily sessions build lasting knowledge for self-directed Learning, exam prep, language study, and Learning a paid skill in 90 days.

This page covers what research supports, how to write review items, how to schedule practice, which free tools to consider, how a personal wiki or notes can feed a deck, and how to start in seven days.

Why forgetting is the real problem

Most learners treat study as intake: highlight a chapter, rewatch a video, or reread notes before a test. That feels productive because the material looks familiar. Familiarity is not the same as recall. When you must produce an answer at work or on an exam, the material often fails to come back.

Forgetting after cramming is normal. Memory research shows that massed practice (many repetitions in one sitting) can help for a short time, yet spaced practice tends to win when the test comes days or weeks later. Ideal gaps between study sessions grow as the desired retention interval grows. A gap that works for a quiz tomorrow is too short for a skill you need in three months.

The practical goal is not perfect memory of everything. It is reliable recall of the facts, terms, procedures, and decisions that matter for your current project, course, or paid skill.

What research shows

Two findings sit under almost every useful spaced-repetition system.

The spacing effect

Spreading study of the same material across time beats packing the same minutes into one block, especially for later tests. A large quantitative review of distributed practice in verbal recall found hundreds of comparisons supporting spacing and lag effects, and showed that the best gap depends on how long you need to remember the material.

The testing effect and retrieval practice

Trying to recall information strengthens memory more than restudying it, especially after a delay. Classic experiments with educational prose found that practice tests beat repeated reading on delayed retention, even when readers felt more confident after rereading. Education-facing summaries of retrieval practice explain the same pattern for classroom and self-study use.

Spaced repetition software combines these ideas: you retrieve an answer, judge difficulty, and the schedule pushes the next review farther out when you succeed and closer when you fail. The method is more than a reminder to reread notes.

How to practice: active recall versus reread

A useful session looks like this:

1. Hide the answer. Cover the page, close the note, or flip the card.
2. Produce. Say, write, or type the answer from memory. For skills, perform the next step of the procedure.
3. Check. Compare with the source. Fix mistakes immediately.
4. Rate difficulty. Easy items wait longer. Hard items return sooner.
5. Stop when the daily queue is done. Mild difficulty during recall supports durable learning, a point often stressed in work on desirable difficulties.

Rereading and highlighting can help when you first meet a topic. They are weak as the main long-term method. If study feels comfortable every minute, you are probably recognizing rather than retrieving. See also Retrieval practice and Learning.

Writing good review items

Bad cards waste months. Good items are short, specific, and demand production. Public knowledge-formulation rules from SuperMemo remain a clear checklist: keep items atomic, ask about one idea at a time, use clear wording, and avoid questions that only make sense in missing context.

Facts and vocabulary. Prefer production over recognition. Example prompt: Name one job TCP handles that UDP does not. Example answer: Reliable ordered delivery with retransmission (among other differences).

Procedures. Break work into steps with cues that mirror real use. Example prompt: After cloning a repo, what command creates and switches to a branch named feature-x? Example answer: git checkout -b feature-x (or the git switch -c form you actually use).

Concepts. Prefer explain or decide prompts over definition dumps. Example prompt: In one sentence, when should a learner prefer short-answer recall over multiple choice? Example answer: When the goal is to practice producing the idea, not only recognizing it among options.

Paid skills. Cards should track tasks clients pay for. Bookkeeping cards might ask how to categorize a transaction type. Spreadsheet cards might ask when to use a formula pattern. Tie the deck to Learning a paid skill in 90 days so practice maps to portfolio pieces, not trivia.

Aim for cards you can answer in a few seconds of thinking. If an item needs a paragraph, split it.

A simple schedule

You can start without software. A paper box or notebook with date columns works:

Review after	Typical use
Same day	First pass after learning (optional quick check)
Day 1	Next-day recall
Day 3	Early consolidation
Day 7	End of first week
Day 14	Two-week check
Day 30	Monthly retention

When you miss an item, move it back to an earlier slot. When you answer easily several times, lengthen the gap.

Apps automate this with scheduling algorithms. They show a daily due queue, update intervals from your ease ratings, and keep hard items near while retiring easy ones for weeks or months. Manual schedules are enough for a few dozen facts. Past a hundred active items, software usually saves time and reduces abandoned decks.

Tools overview

No tool replaces clear items and daily retrieval. Choose by cost, control, and sync needs.

Approach	Cost	Strengths	Tradeoffs
Paper cards or a date box	Free	No setup, works offline, teaches the method	Hard to manage hundreds of cards; no sync
Anki	Free on desktop and AnkiWeb; mobile options vary	Strong scheduling, large community, media cards	Learning curve; easy to overbuild decks
Mnemosyne	Free	Open source, research-oriented scheduling	Smaller ecosystem than Anki
Notebook plus calendar reminders	Free	Fits people who already keep notes	Easy to slip into rereading without self-tests

Pick one system and stay with it for at least two weeks before switching. Changing tools often is a common way to avoid writing good prompts.

Personal wiki, notes, and optional AI card drafts

Spaced repetition works best when cards come from material you already care about. A personal notes system or wiki can be the source of truth: pages hold explanations, examples, and sources; cards hold only the atomic prompts you must retrieve. See How to run a personal wiki and Personal knowledge management.

A simple pipeline:

1. Capture. After a study session, list five things you must still be able to do next week.
2. Distill. Turn each into one clear question and one short answer in your notes.
3. Import. Move verified pairs into your deck or paper box.
4. Link back. Keep a pointer from the card topic to the longer page when an item keeps failing.

Optional AI help: a local or cloud model can draft candidate cards from a paragraph you paste. Treat drafts as raw material. Verify every fact, delete vague items, and keep only prompts you can answer without the model. This pairs with careful use of Local AI as a public library style tools. Do not paste private client data into a cloud tool without permission. Do not let generated cards replace checking the original source. You remain responsible for accuracy.

Common mistakes

• Too many new cards. Adding dozens on day one creates a review debt you will abandon. Cap new cards (for example 5 to 15 per day) until reviews feel sustainable.
• Passive recognition. Peeking at the answer or relying only on multiple choice trains recognition, not recall.
• Giant multi-idea cards. If the answer is a wall of text, split it.
• Decks without a purpose. A deck for a paid skill or a live course beats a random trivia pile.
• Skipping the daily queue. Missing several days turns the queue into a pile. Do a short catch-up rather than deleting the deck.
• Confusing tools with practice. Installing software is not studying. Opening the due list and retrieving is studying.
• Never retiring topics. When a project ends, suspend unused cards so the daily queue stays tied to current goals.

Making the daily review visible (streaks, a simple score for sessions completed) can help consistency; see Making productive work as engaging as games.

7-day starter plan

Use this if you are starting from zero. Even 20 focused minutes a day is enough to test the method.

• Day 1: Choose one target. Pick a single course unit, language chapter, or paid-skill topic. Write five must-know outcomes. Create five atomic cards. Review them once after a short break.
• Day 2: First spaced pass. Review yesterday without looking first. Add up to five new cards from the same topic. Log misses.
• Day 3: Fix card quality. Rewrite confusing cards. Split oversized items. Do the full due set.
• Day 4: Connect to notes. Move explanations into a personal wiki or notes page. Keep cards short. Add five cards that test decisions, not only definitions.
• Day 5: Skill application. Do a real practice task. Add cards for mistakes you actually made.
• Day 6: Optional AI draft, human verify. Ask a tool to propose ten cards from one notes page. Keep at most five after you check sources. Review the due queue.
• Day 7: Review the system. Count cards, minutes, and retention on a short self-quiz from scratch. Set your daily new-card cap and your tool for the next three weeks. Plan the day-14 and day-30 checks on a calendar.

After week one, keep the daily due queue sacred. Add new material only at a rate your future self can review.

See also

• Learning
• Online learning
• Teaching
• Retrieval practice
• How to run a personal wiki
• Personal knowledge management
• Learning a paid skill in 90 days
• Making productive work as engaging as games
• Local AI as a public library
• Areas of research and study to develop

Discussion questions, essay ideas, and learning related AI prompt ideas

• Why does rereading feel effective even when delayed tests favor retrieval practice?
• How should a learner choose gaps between reviews when preparing for a test in two weeks versus a skill needed in six months?
• Essay idea: Compare a week of cramming with a week of spaced recall for the same material and describe what each produces.
• Essay idea: Design a card set for one paid skill and explain how each card maps to client-facing work.
• Essay idea: Argue when a personal wiki should hold full notes and when a spaced-repetition deck should hold only prompts.
• AI prompt idea: Make ten short-answer review questions from the notes I paste and flag any question that needs a source check.
• AI prompt idea: Turn this procedure into five atomic step cards with clear prompts and short answers.
• AI prompt idea: Quiz me on the topic from yesterday with increasing difficulty and tell me which answers were incomplete.
• AI prompt idea: Build a four week spaced review calendar for these twenty items with day 1, 3, 7, 14, and 30 checkpoints.