What spaced repetition means
Return to material in separate sessions over time. This can involve cards, problems or open questions. Adaptive scheduling is one way to organize it, but spacing does not require an app or a particular algorithm.
There is no universal forgetting percentage or exact review time for everyone. Difficulty, familiarity and the period you want to retain the material influence the schedule.
What research supports
Cepeda and colleagues’ 2006 meta-analysis, linked below, examines distributed practice in verbal recall tasks. Useful intervals also depend on the delay before the final test. It does not establish that an app can detect an exact moment before forgetting.
An example you can adapt
Choose a small topic, answer questions today and try again in a later session. If an answer is easy to recall, consider a longer gap; if you miss it, revisit the explanation and return sooner. This illustrates a decision, not a prescribed schedule.
Record your result before looking at the answer. Recognizing a sentence after seeing it does not establish that you could recall it.
- Record the topic, date and whether the answer was complete.
- Investigate mistakes before adding repetitions.
- Reduce new material if pending review becomes unmanageable.
Use other practice when the task demands it
Cards support specific questions. For writing, proving a theorem or solving a case, add tasks requiring those skills. Knowing a formula is different from knowing when to use it.
In Lumia AI, organize cards and use the available review workflow. Assess answers honestly and correct generated content before studying it.
Choose intervals for the period you need to remember
Remembering a topic for an upcoming test is different from keeping it available throughout a course. Define the date or period that matters and observe whether you can retrieve the answer after a real gap. When the goal changes, a reasonable practice frequency may change as well.
An early miss does not require restarting the entire deck. Check whether knowledge was missing, the card was ambiguous or the answer demanded too many ideas at once. Fix the cause first; repeating a defective question more often preserves the defect.
Audit the schedule with a manageable sample
Before applying one schedule to an entire course, choose a sample of questions from a single topic and set a date when you want to check them again. Record each attempt before revealing the answer, the elapsed interval and the type of error. Keep the same rule for deciding whether an answer is complete; otherwise, you will be comparing different scoring decisions.
After several sessions, review which questions remain useful, which need clearer wording and how much review fits your schedule. If pending reviews keep growing, temporarily reduce new material. This audit can help adjust one workflow; it does not measure your general memory capacity or show that the schedule alone caused every change you observed.
- Set a target date and a defined check task.
- Separate knowledge, wording and attention errors.
- Fix or remove defective cards before changing intervals.
- Keep applied practice outside the deck.
Common mistakes in spaced review
Adding new material without a limit can create a review queue that no longer fits your schedule. Set a load you can complete and pause additions when pending work exceeds it. Avoid rating an answer as easy after seeing it; record what you recalled before revealing the solution.
An app schedule is an operational recommendation, not a direct measurement of memory. Check learning with unfamiliar questions and application tasks. The linked Dunlosky review and Cepeda meta-analysis support distributing practice, but they do not validate one interval for every subject or person. A simple dated log can help you inspect the same decisions without relying on hidden scheduling assumptions.
- Do not copy another person's schedule without considering the task and deadline.
- Do not treat recognition after revealing an answer as independent recall.
- Do not preserve outdated cards simply to retain their history.
- Do not replace problems, writing or oral practice with short questions.
Sources and scope of this guide
These references provide context on studying and learning. Examples are illustrative; they are not an evaluation of Lumia AI and do not guarantee grades or retention rates.
- Cepeda et al. (2006): Distributed practice in verbal recall tasks — Meta-analysis of distributed-practice findings in verbal recall tasks.
- Dunlosky et al. (2013): Improving Students’ Learning With Effective Learning Techniques — Comparative review of ten learning techniques and the conditions in which they were studied.
Editorial approach: the body names the source beside evidence-based claims; this list identifies the document, link, and scope consulted. We distinguish those references from our own examples. If you find an error, include the page and sentence when you contact the team.
Apply spaced repetition to your own materials
Build a reviewed deck and treat suggested intervals as guidance you can adjust.
Get started free