Choose a note-taking method from two things: how the material is structured, and how you intend to review. Cornell notes suit lectures you will self-test from, outlining suits hierarchical material, mapping suits relationships between concepts, charting suits comparable categories, and sentence notes suit fast unstructured input you will rewrite later.
Creator disclosure: Heni Hazbay creates Summarise Visually and has a direct interest in the app. The study guidance below comes from published sources, not from the app.
Publication and evidence note: This guide was prepared from the cited sources and reviewed project evidence. Its comparison table, selection test, and worked example are editorial tools, not a ranking validated by a controlled comparison of methods.
The five methods at a glance
| Method | Shape on the page | Best for | Main cost | Fails when |
|---|---|---|---|---|
| Cornell | Cue column, note column, summary line | Lectures and readings you will self-test | Set-up and a post-session pass | You skip the cue column and it becomes plain notes |
| Outline | Nested headings and indents | Hierarchical, well-organised material | Requires structure to exist | The speaker jumps around and the nesting breaks |
| Mapping | Nodes and links radiating from a central idea | Relationships, causes, systems | Slow to draw live | Detail matters more than relationships |
| Charting | Table with fixed columns | Comparable items across the same dimensions | Columns must be decided in advance | An item does not fit the columns you chose |
| Sentence | One numbered line per point | Fast, unpredictable input | Almost no structure captured | You never do the rewrite pass |
Cornell: notes designed to be tested
Cornell University’s system divides the page into a narrow cue column, a wider note-taking column, and a summary strip at the bottom. During the session you take notes on the right. Afterwards you write questions or cues on the left and a short summary underneath. To review, you cover the note column and answer the cues.
That structure is the reason it is worth the set-up: it produces prompts you can attempt without seeing the answers. Dunlosky and colleagues’ review rated practice testing as a high-utility technique, and the cue column is a practice-testing format built into the notes themselves.
The common failure is skipping the post-session pass. Without cues and a summary, Cornell notes are just notes in a narrow column.
Outline: hierarchy you can trust
Outlining records main points with sub-points indented beneath them. It is fast, familiar, and produces material that converts cleanly into revision headings.
It depends on the source having a hierarchy. A well-organised textbook chapter or a structured lecture outlines easily. A discussion-led seminar or an interview does not, and forcing an outline onto unstructured input produces indentation that implies relationships nobody stated.
Mapping: relationships over detail
Mapping places the central concept in the middle and branches outward, labelling the links. It is the right choice when the thing to learn is how parts relate — a mechanism, a causal chain, a system with feedback.
Maps also lend themselves to dual coding: pairing words with a visual representation of the same content. The Learning Scientists describe dual coding as combining verbal material with visuals so the same idea has two complementary representations. Note that this means a diagram of the content, not decoration; an image that carries no information adds nothing to retrieve.
Mapping live is slow. Most people map after the session, from rougher notes.
Charting: when items are comparable
Charting sets up columns before the session — for example theory, proponent, evidence, criticism — and fills rows as material arrives. When the material genuinely is a set of comparable items, this is the most efficient method available, and the finished chart is already a revision object.
Its weakness is rigidity. If an item does not fit your columns, you will either distort it or drop it. Leave a “does not fit” row and use it honestly.
Sentence: capture now, structure later
Sentence notes are one numbered line per point, in the order they arrive. They are the fallback for fast, unpredictable input where any structure you impose live would be wrong.
They are only useful if you do the second pass: within a day, convert them into one of the structured methods above. Left alone, they are a transcript of your attention rather than a study object. The distinction between raw capture and usable notes is covered in lecture transcript versus lecture notes.
A two-question selection test
- What shape is the material? Hierarchical, relational, comparable, or unpredictable? That eliminates two or three methods immediately.
- How will I review? If the answer is self-testing, choose a method that produces prompts — Cornell, or charting where the row headers become the questions. If the answer is “reread,” change the review plan rather than the method; rereading was rated low-utility in Dunlosky’s review. How to remember what you read covers the review plan itself.
A worked example
Consider a fictional module with three sessions. This example is illustrative.
Session one is a structured lecture introducing four theories with named proponents and standard criticisms. Comparable items across identical dimensions: charting, with the columns set up beforehand.
Session two is a seminar where students debate one case and the discussion doubles back repeatedly. Unpredictable input: sentence notes live, rewritten into Cornell format the same evening, with the cue column drawn from the questions the debate kept returning to.
Session three is a reading on a feedback mechanism where the parts only make sense together. Relationships: mapping after a first read, with the links labelled rather than left as bare lines.
The module used three methods because it contained three shapes of material. Using one method throughout would have cost more than the switching did.
Where a summary tool fits
Current project review supports six summary modes, including Key Points and Q&A, applied to supplied text such as a transcript or pasted passage. A generated Key Points list is closest to sentence notes: a flat capture that still needs structuring. A generated Q&A set resembles a cue column, but the questions come from the text as supplied rather than from your syllabus or your gaps.
So the honest placement is this: use a summary to compress raw material you have permission to use, then apply the method yourself, then test yourself from your own cues. The lecture summary template and textbook chapter notes template give you the structured records for each source type, and key points and Q&A for study covers turning them into a retrieval loop.
Frequently asked questions
Which note-taking method is best for lectures?
For a structured lecture you will be examined on, Cornell notes are a strong default because they build prompts for self-testing. For a lecture comparing several items, charting is usually more efficient.
What is the Cornell method in one sentence?
Divide the page into a cue column, a note column, and a summary strip; take notes during the session, write cues and a summary afterwards, then cover the notes and answer the cues.
Is handwriting better than typing?
Evidence on this is mixed and depends on what is measured and how people use each medium. The more reliable difference is what you do afterwards: notes you self-test from outperform notes you only reread, in either medium.
Can I use more than one method for one course?
Yes, and you usually should. Match the method to the shape of each session’s material rather than to habit.
Do mind maps actually help?
They help when relationships are what you need to learn, and when the visual carries content rather than decoration. For dense factual detail, a structured list or chart is usually easier to test from.
Source and evidence note
The cue column, note column, and summary structure comes from Cornell University’s note-taking system. The relative utility of practice testing and rereading comes from Dunlosky and colleagues’ review in Psychological Science in the Public Interest. The description of dual coding as pairing verbal material with informative visuals comes from The Learning Scientists. All three were accessed on July 25, 2026.
The comparison table, the two-question selection test, and the fictional three-session module were created for this guide. They are editorial tools; no controlled comparison of these methods was conducted, and no method is claimed to be superior in general. Product statements reflect current project review.
Limits and verification
Method choice is less important than what happens after the session. Notes of any shape that are never retrieved from will not support recall, and a method that suits your material can still fail if the review plan is rereading.
AI assistance can restructure captured text quickly, but it can also impose a hierarchy the source never stated or generate questions about points the source treats as minor. Check generated notes against the original material with the AI summary accuracy checks, record only material you have permission to use, and follow your institution’s rules on recording and AI use.
Publication was authorized by Heni Hazbay after review of the cited sources and current project evidence.