Successive Relearning for YouTube Lectures (Retrieve Until Correct, Then Come Back Later)

Published 2026-10-23 · Francesco Masciopinto

Most students already know two pieces of advice: test yourself, and space your review. Successive relearning is what happens when you combine them into one simple rule. You quiz yourself on a lecture until you can answer every question correctly, then you come back in a few days and do it again, still until every answer is correct. Each later session is usually shorter, and that shrinking time is your evidence that the material is sticking.

This guide shows how to run successive relearning for YouTube lectures: how to turn a video into questions, what "correct" means, how to schedule the return sessions, and how to stop the routine from eating your whole week.

What successive relearning actually is

The idea comes from learning-science research on retrieval practice. Instead of reviewing a fixed number of times, you use a criterion: keep retrieving until you hit it. Then you relearn in later sessions, again to criterion.

Two parts make it different from casual flashcard review:

Initial learning to criterion. In the first session you do not stop after one pass. You cycle through the questions, setting aside each one you answer correctly, and keep going with the missed ones until each has been answered correctly at least once (many students use one correct answer for easy items and two for hard ones).
Relearning sessions. Days later you start over with the full set and again work to criterion. Items you forgot come back quickly because you are relearning, not learning from zero.

You are not trying to feel fluent. You are trying to produce the answer from memory, repeatedly, across days.

Why it suits lecture videos

YouTube lectures have a built-in weakness: they are easy to rewatch. Rewatching feels like studying, but recognition ("oh right, I remember this slide") is not recall. Successive relearning forces the opposite. The video becomes a source you consult only after you fail to retrieve, and the timestamp tells you exactly where to look.

It also works well with long playlists. Each lecture becomes a small, finite question set, so you always know what "done for today" means.

Step 1: Turn one lecture into a question set

After watching (or during a second, faster pass), write 10–20 questions per hour of lecture. Good questions have one checkable answer and point back to a timestamp.

Question type	Example	Why it works
Definition	What does "opportunity cost" mean in the lecturer's words?	Anchors vocabulary the exam will use
Mechanism	What are the three steps that lead from X to Y?	Forces ordered recall, not just a term
Contrast	How is A different from B, and when does it matter?	Targets the classic exam trap
Application	Given this small scenario, which rule applies?	Checks transfer, not memorised wording
Worked step	What is the next step after setting up the equation?	Rebuilds procedures from memory

Keep each answer short enough to check in a few seconds. If an answer needs a paragraph, split it into two or three questions.

Step 2: Use a simple template

You can keep this in a notes app, a spreadsheet, or paper. What matters is the timestamp and the session log.

LECTURE: [course] — L07 [title]            VIDEO: [link]
CRITERION: 1 correct (easy) / 2 correct (hard items marked *)

#   Question                                   Answer (short)          Time
1   Define ____ as the lecturer did            ...                     04:12
2*  Three steps from ____ to ____              a → b → c               11:30
3   ____ vs ____: when does it matter?         ...                     18:05
...

SESSION LOG
S1  2026-__-__   rounds: 4   minutes: 28   missed first try: 9/16
S2  +2 days      rounds: 2   minutes: 12   missed first try: 4/16
S3  +7 days      rounds: 1   minutes: 7    missed first try: 1/16

The session log is the motivating part. Watching minutes and misses fall is visible proof that review is working.

Step 3: Run the first session to criterion

Shuffle the questions or start from the top.
Answer each out loud or in writing before checking.
Correct? Mark it and set it aside for this session (or until its second correct answer if it is a starred hard item).
Wrong or blank? Check the answer. If you do not understand the answer, jump to the timestamp and watch only that segment, usually 1–3 minutes.
Loop through the remaining questions until every one has reached criterion.

Do not rewatch the whole lecture in this session. The point is that the video serves retrieval, not the other way round.

Step 4: Schedule relearning sessions

There is no single magic interval. A practical schedule for a semester course:

Session	When	Typical length	Goal
S1	Same day or next day after watching	20–40 min	Reach criterion for the first time
S2	2–3 days later	10–20 min	Recover what faded fastest
S3	About 1 week later	5–15 min	Confirm stable items
S4	2–3 weeks later or before the exam	5–10 min	Final check; rewrite anything still failing

If a session still takes as long as the first one, your questions may be too vague or too large. Rewrite them rather than adding more sessions.

Step 5: Fit it into a weekly lecture load

If you watch three lectures a week, relearning sessions stack up. Keep them small:

Cap each lecture at 20 questions. More than that and S2 becomes a chore.
Batch relearning. Do S2 for Monday's lecture and S3 for last week's in one 30-minute block.
Retire stable items. If a question has been correct on first try in two relearning sessions, move it to a "check before exam" list.
Protect new learning. Do not let old relearning push out watching the next lecture. Two short blocks beat one exhausted block.

Making better questions over time

Your first question sets will be uneven. After S2, look at which questions you missed:

Missed because the question was ambiguous? Rewrite it.
Missed because two ideas blur together? Add a contrast question.
Missed because a step is always forgotten? Split the procedure into smaller questions.
Never missed? It may be too easy; retire it early.

Successive relearning is also a diagnostic. The items that keep failing in S3 are the ones to take to office hours or a study group.

Common mistakes

Stopping at one pass. Without a criterion you are doing ordinary review, not relearning.
Checking the answer before trying. Even a wrong guess helps more than reading the answer immediately.
Rewatching instead of consulting. Use the timestamp for the specific gap, not the whole video.
Questions that are paragraphs. Long answers cannot be checked quickly and quietly turn into rereading.
Skipping S2. The first relearning session is often where most forgetting shows up; it is also usually short.

Find the right segment faster

When a missed question sends you back to the video, the slow part is finding the exact minute. SummarizAI is a Chrome extension that puts a summary, chapters, chat, and Study flashcards on the YouTube watch page, which helps you draft first-pass questions and jump to the relevant segment. The free plan works as a student trial. The relearning itself, answering from memory until you get it right, still has to be done by you.

Frequently asked questions

How is successive relearning different from spaced repetition apps?

Spaced repetition apps schedule individual cards with an algorithm. Successive relearning is a session rule: retrieve to criterion, then relearn the whole set later. You can run it inside an app or on paper; the criterion is what matters.

What criterion should I use?

One correct answer per session is a sensible default. For items that keep failing or that the exam weights heavily, require two correct answers in the same session.

Do I need to watch the lecture before making questions?

Yes, at least once. Questions written from a summary alone can be useful for planning, but you should confirm each answer against what the lecturer actually said.

How many relearning sessions are enough?

For most courses three or four sessions per lecture, spread over a few weeks, are realistic. Stop adding sessions when items are correct on first try and the session takes only a few minutes.

Can I use this for problem-solving courses?

Yes. Turn worked examples into "what is the next step?" questions, and add one full problem per lecture that you solve from a blank page in each session.

Related guides

Try SummarizAI on your next lecture

SummarizAI is a Chrome extension that adds a summary, chapters, chat, and Study flashcards on the YouTube watch page. The free plan is a student trial.

Try SummarizAI free

Students · Study from lectures · Lecture notes · Install