WHY ENGLISH?
A literature map can reveal a research neighbourhood; careful reading decides which doors are relevant, reliable and worth opening.
This guide shows students, parents and educators how to build, read and maintain a Litmaps project while preserving source identity, search limits and human judgement.
Litmaps turns scholarly records into visual literature maps built from starting articles, citation connections and other recommendation signals. It can help a learner find related papers, organise a literature review and monitor new work. English matters because terms such as related, connected, similar, important and gap are not interchangeable.
Students searching for a literature review tool, citation map, related research papers or new-paper alerts often hope the interface will tell them which sources are best. It cannot. A map is a discovery representation. The article remains the object that must be identified, read and evaluated. Search settings, database coverage and the chosen seed articles shape what appears.
The practical value of English lies in calibrated sentences: “the map suggested”, “the record lists”, “the paper reports”, “the method supports” and “the search did not identify”. These verbs let a student enjoy visual discovery without turning a network into a verdict. The same discipline supports multilingual research by making language and database limits visible.
Begin With a Research Question, Not a Pretty Map
A visual map feels productive immediately, but a research question determines whether the activity is purposeful. Write the population, concept, outcome, setting and period where they matter. A broad exploratory question can be appropriate; it should still be explicit.
The question controls seed selection, search terms, screening rules and stopping decisions. Without it, every neighbouring paper can look relevant. With it, a student can explain why an article enters the map and what role it may play.
A practical example: Begin With a Research Question, Not a Pretty Map
For “How does retrieval practice affect delayed vocabulary learning among secondary students?”, a paper about university anatomy may offer a mechanism but not direct population evidence. It can be tagged as background rather than silently treated as an answer.
A map can help refine a question. Revision is healthy, but record the old and new versions so that search decisions remain understandable.
Next steps: Begin With a Research Question, Not a Pretty Map
Write a one-sentence question and three exclusion rules before opening the mapping tool.
Choose Seed Articles as Lenses
Litmaps calls the starting records input or seed articles. The official Create a Litmap guide says users can search for or upload starting articles and then explore related work. Every seed is a lens: its references, citations, authors, vocabulary and age shape the neighbourhood.
Use more than popularity. A recent review can offer breadth, a close empirical study can offer specificity and a contrasting approach can reduce one-seed tunnel vision. Verify title, authors, year, venue and identifier before building.
A practical example: Choose Seed Articles as Lenses
A student starts with a systematic review, a Singapore classroom study and a foundational experiment. Papers recurring across several maps deserve inspection, while unique papers may reveal a specialised branch.
Recurrence is not proof of quality or independence. Several articles can belong to one study family or repeatedly cite the same assumptions.
Next steps: Choose Seed Articles as Lenses
Record why each seed was chosen and which version it represents. Replace an accidental seed when identity or scope proves wrong.
Understand Citation Links Before Interpreting Distance
The official Litmaps documentation says its default Shared Citations and References search works from citation networks. A citation relationship is a documented link between scholarly records; it is not automatically agreement, replication or endorsement.
Read direct references, later citations and co-citation patterns as different relationships. A paper can cite another to criticise it. Two papers can be cited together because later authors compare opposing methods.
A practical example: Understand Citation Links Before Interpreting Distance
If two nodes sit near each other, write first: “Litmaps placed these records together under the selected search and view.” Only after reading may the student say they share a method, theory or conclusion.
Screen geometry is not a probability of truth. Layout, filters and algorithm settings can change position.
Next steps: Understand Citation Links Before Interpreting Distance
Translate every visual observation into a bounded sentence, then list the source text needed to support any stronger claim.
Read the Search Algorithm Label
Litmaps currently documents three search algorithms: Shared Citations and References, Common Authors and Similar Text. Each answers a different discovery question. Algorithm names belong in research notes because results can change when the mechanism changes.
Citation-based discovery can find connected work that uses different keywords. Common Authors can reveal research teams but may reinforce established networks. Similar Text uses titles and abstracts, helping surface semantically close work that may have few citations.
A practical example: Read the Search Algorithm Label
A new paper with limited citations may appear under Similar Text but not under a citation-heavy search. That difference does not make one result wrong; it reveals what each method can see.
Search descriptions and features can change. Verify current official documentation instead of copying an old tutorial.
Next steps: Read the Search Algorithm Label
Run at least two relevant algorithms for an important topic and record which papers appear only under one approach.
Treat Database Coverage as a Boundary
Litmaps says its database combines open metadata from providers including Crossref, Semantic Scholar and OpenAlex, while acknowledging that records can be absent when metadata are unavailable. Large coverage is not complete coverage.
Coverage varies across disciplines, document types, languages and time. Books, local reports, non-English sources, very recent articles and incomplete records may be underrepresented. A literature-map search should complement library databases, subject indexes and expert reference chasing.
A practical example: Treat Database Coverage as a Boundary
A Singapore education question may require MOE publications, institutional repositories and local theses that are not prominent in a global citation network.
Do not convert a zero-result map into “no research exists”. The defensible sentence names the system, settings and date checked.
Next steps: Treat Database Coverage as a Boundary
Keep a search log with databases, date, language, query or seeds, filters and result count. Use gaps in coverage to choose the next source.
Read Nodes as Records, Not as Papers You Have Read
A node points to a scholarly record. Until the learner opens the source, the node offers metadata and network position, not knowledge of method or result.
Check title, ordered authors, year, DOI, document type, abstract and access route. Resolve preprint, conference and journal versions. Use the publisher or repository record for formal status rather than inferring it from map appearance.
A practical example: Read Nodes as Records, Not as Papers You Have Read
A map contains two similar titles by the same authors. One is a conference paper and one a later journal article. Merging them in notes would double-count the study and hide revision.
Metadata can be incomplete or wrong. Citation counts change and should not be treated as fixed quality scores.
Next steps: Read Nodes as Records, Not as Papers You Have Read
Before tagging a node “evidence”, open the record and add a version-aware identity note.
Use Map Views to Ask Different Questions
Current Litmaps documentation describes citation, semantic, co-citation and co-author views, with some views or grouping features tied to paid access. These layouts organise records by different relationships.
A citation view can help trace time and linkage. A semantic view can cluster similar titles and abstracts. Co-citation can reveal works used together, while co-author networks can show research teams. Group labels may be machine-generated suggestions and require inspection.
A practical example: Use Map Views to Ask Different Questions
A cluster labelled “student engagement” contains papers using attendance, self-report and platform clicks as different measures. The label is a starting hypothesis, not a shared operational definition.
Feature availability and pricing can change. An article should explain the mechanism without promising access on every account.
Next steps: Use Map Views to Ask Different Questions
Open three papers from one group and write the exact commonality plus one important difference before accepting the group name.
Notice Bridges, Clusters and Outliers Carefully
Visual networks help readers notice structure. A bridge node may connect methods or subfields. A dense cluster may share citations. An outlier may signal a novel approach, incomplete metadata or a mistaken seed.
Use these shapes to generate questions: What vocabulary connects the groups? Is the bridge a review, method paper or interdisciplinary application? Does the outlier truly belong? Then inspect sources.
A practical example: Notice Bridges, Clusters and Outliers Carefully
A bridge paper links cognitive psychology and classroom practice. Reading shows that it adapts a laboratory task for school use. That verified mechanism belongs in the literature review; the bridge position alone does not.
Cluster size is influenced by database coverage and search settings. Small fields and regional work may appear peripheral despite high relevance.
Next steps: Notice Bridges, Clusters and Outliers Carefully
For every visual pattern, write two alternative explanations and the evidence needed to distinguish them.
Build Tags Around Decisions, Not Decoration
Litmaps lets users save articles with tags. Tags become useful when they answer future questions: why was this saved, what should happen next and what role might it play?
Use a small controlled vocabulary such as include, exclude-reason, background, method, close-population, counterevidence, verify-version and unread. Keep topic tags separate from workflow tags.
A practical example: Build Tags Around Decisions, Not Decoration
A paper tagged “include” also carries “method: quasi-experimental”, “population: secondary” and “next: check attrition”. Another carries “exclude: wrong outcome”. The library remembers reasoning, not just enthusiasm.
Too many synonyms create retrieval noise. A tag called “important” hides the criterion.
Next steps: Build Tags Around Decisions, Not Decoration
Write a tag dictionary with definitions and examples. Review unused or ambiguous tags before collaboration.
A Worked Example: Map Classroom Feedback Research
Imagine a JC student examining whether written teacher feedback improves revision quality. The student begins with a recent review, a classroom intervention and a paper on feedback literacy.
The map reveals clusters around corrective feedback, formative assessment, peer feedback and student uptake. The learner tags possible background, direct evidence, mechanism and counterevidence, then opens representative sources.
A practical example: A Worked Example: Map Classroom Feedback Research
One large node is a broad review but measures attainment rather than revision quality. A smaller classroom study matches the outcome but has a limited sample. The final synthesis describes both roles instead of ranking by node size.
The map does not supply an exhaustive systematic review. Language, database and date limits remain visible.
Next steps: A Worked Example: Map Classroom Feedback Research
Create an evidence table with question, population, design, feedback type, outcome, follow-up, finding and limitation. Cite papers, not the map.
Monitor Alerts Need a Screening Rule
Litmaps Monitor can rerun a map search and send alerts when new articles are added. Official documentation says all users can enable Monitor, while custom frequency, algorithm and filters are tied to Pro features at the time checked.
An alert is a candidate, not an update to the conclusion. Screen title and abstract, verify version, assess eligibility and record a decision. New records can be old papers newly indexed.
A practical example: Monitor Alerts Need a Screening Rule
A monthly alert surfaces a 2022 paper. The learner writes “newly found” rather than “new research” and checks why it was missing from the earlier map.
Email frequency and feature access may change. More alerts can increase noise and confirmation bias.
Next steps: Monitor Alerts Need a Screening Rule
Define include, exclude and maybe rules before enabling Monitor. Review alerts on a schedule rather than citing directly from email.
Verify Dates, Versions and Corrections Outside the Map
A map record can point toward an article, but version history and corrections may live on publisher, repository, DOI or retraction pages.
Compare preprint, accepted manuscript and version of record. Check correction, expression-of-concern or retraction notices. Match figures and conclusions when using an accessible manuscript.
A practical example: Verify Dates, Versions and Corrections Outside the Map
A preprint node and journal node appear separately. The journal version changes one analysis and adds a limitation. The student keeps the relationship and uses the correct version for the claim.
The newest date may be online publication, issue assignment or index update. State the date type when chronology matters.
Next steps: Verify Dates, Versions and Corrections Outside the Map
For central sources, follow DOI and publisher records, then use the eduKate guide to reading retraction and notice records.
Do Not Call a Sparse Area a Research Gap Too Quickly
A blank space on a map may reflect a genuine under-studied question, but it may also reflect missing metadata, wrong seeds, language limits, terminology changes or weak citation links.
A research gap requires evidence across searches and sources. Distinguish no studies found, few studies found, inconsistent evidence, missing population, missing outcome and weak method. These are different gaps.
A practical example: Do Not Call a Sparse Area a Research Gap Too Quickly
A map shows little research on Singapore lower-secondary students. Local repositories reveal several relevant theses using different terminology. The initial visual gap was partly a vocabulary and coverage gap.
Claiming novelty can miscredit prior researchers and weaken a proposal.
Next steps: Do Not Call a Sparse Area a Research Gap Too Quickly
Search alternative terms, languages, repositories and reference lists. Write the narrowest gap statement supported by the search record.
Keep Mapping Separate From Systematic Searching
A visual tool can support scoping and citation chasing, but formal systematic reviews require protocol-driven searches, multiple databases, reproducible strategies and transparent screening.
Use Litmaps to discover vocabulary, influential records, citation chains and new papers. Do not describe the result as exhaustive unless the review method and coverage justify that claim.
A practical example: Keep Mapping Separate From Systematic Searching
A school project may appropriately say “We used Litmaps and database searches to identify illustrative sources.” A dissertation review may need a registered protocol and database-specific search strings.
The right standard depends on the task. Overclaiming rigour is not an upgrade.
Next steps: Keep Mapping Separate From Systematic Searching
Ask the teacher, supervisor or librarian what review type is expected, then name the method accurately.
Write Relationship Sentences After Reading
A literature review needs verbs that express verified relationships: defines, adapts, tests, extends, replicates, qualifies, contradicts, synthesises and reanalyses.
Map proximity supplies none of these verbs by itself. Read the relevant sections and record page, table or figure support. Distinguish author claims from your interpretation.
A practical example: Write Relationship Sentences After Reading
Instead of “These papers are connected,” write: “Tan et al. adapt the feedback rubric introduced by Lee, while Kumar tests a related rubric with a different age group.”
Strong verbs can still overclaim if the study design is weak. “Causes” requires more than chronological sequence or association.
Next steps: Write Relationship Sentences After Reading
Build a relationship-verb bank and attach one evidence note to every sentence connecting two sources.
Collaborate Through Shared Workspaces and Decision Logs
Litmaps workspaces can separate projects, maps and tags. Collaboration works only when team members share definitions and record why records move between states.
Assign roles such as seed verifier, screener, full-text reader and conflict resolver. Keep a decision log outside the visual map if the platform does not capture the detail needed.
A practical example: Collaborate Through Shared Workspaces and Decision Logs
Two students disagree about a paper. One sees a matching title; the other sees the wrong outcome. They record both reasons and apply the written inclusion rule instead of voting on appearance.
Shared access can expose notes or copyrighted files. Check permissions and institutional rules.
Next steps: Collaborate Through Shared Workspaces and Decision Logs
Agree on tag definitions, identity checks and conflict resolution before collecting dozens of records.
Export Records Without Assuming the Bibliography Is Correct
Litmaps supports exports such as BibTeX, RIS and CSV according to its help centre. Export is a transport step, not a metadata guarantee.
After import into a reference manager, check authors, title case, year, container, DOI, pages and document type. Deduplicate versions and preserve notes about why each source was included.
A practical example: Export Records Without Assuming the Bibliography Is Correct
An RIS export labels an online-first article with incomplete issue data. The student verifies the DOI record before final citation.
Citation styles cannot repair wrong identity. A beautiful bibliography can still point to the wrong version.
Next steps: Export Records Without Assuming the Bibliography Is Correct
Audit every cited source against an authoritative record and the version actually read.
Use English to Make Search Limits Visible
Multilingual and regional research can be hidden when seed titles and abstracts are predominantly English. Search tools can inherit those visibility patterns.
Record languages searched, translations used and databases not covered. Preserve original titles and names. Seek qualified help where machine translation cannot support technical claims.
A practical example: Use English to Make Search Limits Visible
A student maps English-language papers on bilingual education, then adds Chinese and Malay search terms in appropriate sources. The final review explains the coverage rather than treating English visibility as global completeness.
English can be the synthesis language without becoming a measure of research worth.
Next steps: Use English to Make Search Limits Visible
Add a limitations paragraph that names language, geography, document type, database and date boundaries.
A Parent and Student Routine for a Manageable Review
Parents can support process without selecting conclusions. Ask the student to explain the question, seed choice, one visual pattern, one verified paper and one known limit.
Use short sessions: map for discovery, read for evidence, record identity, then stop. Separate collecting from writing so that a growing graph does not become procrastination.
A practical example: A Parent and Student Routine for a Manageable Review
A student saves five candidates, fully checks two and explains why three remain unread. That is stronger than claiming to have “researched 50 papers” by glancing at nodes.
More sources are not automatically better. The assignment’s depth and time matter.
Next steps: A Parent and Student Routine for a Manageable Review
Set a source budget and stopping rule, then require one synthesis sentence for every two or three sources retained.
Finish With a Reproducible Map Note
A future reader should understand how the map was built even if the interface later changes.
Record research question, seeds and identifiers, date, algorithm, filters, view, tags, number screened, inclusion decisions and supplementary searches. Save exports where permitted.
A practical example: Finish With a Reproducible Map Note
The note says: “Map created 4 November 2026 from three verified seeds; Shared Citations and References plus Similar Text; publications 2018–2026; English records screened; local repository search added separately.”
Reproducibility does not mean another user will see an identical live map after the database updates. It means your decisions are traceable.
Next steps: Finish With a Reproducible Map Note
Attach the note to the project folder and update it when Monitor alerts or new seeds materially change the review.
Compare Keyword Search With Citation Discovery
Keyword search and citation discovery reveal different parts of a field. Keywords depend on the words recorded in titles, abstracts and indexes; citation networks can connect papers that use different vocabulary.
Begin with both routes. Extract synonyms and older terminology from strong papers, then return to databases with improved queries. Use citation discovery to follow relationships around known work.
A practical example: Compare Keyword Search With Citation Discovery
A paper on “testing effects” is relevant to a search for “retrieval practice” even though the exact modern phrase is absent. The map helps recover it; the database query confirms broader coverage.
Citation discovery can favour older, well-linked work, while keyword search can surface isolated or new records. Neither route is neutral.
Next steps: Compare Keyword Search With Citation Discovery
Keep separate counts for keyword-found, map-found and reference-chased sources. Compare what each route contributes to the final evidence set.
Screen Titles and Abstracts With Written Criteria
A visual map can tempt readers to save by colour, size or proximity. Written inclusion and exclusion criteria move the decision back to the research question.
Define population, topic, outcome, method, document type, date and language rules where relevant. Record an exclusion reason rather than deleting a paper silently.
A practical example: Screen Titles and Abstracts With Written Criteria
A closely placed paper studies adult professional learners, while the question concerns lower-secondary pupils. It is excluded from direct evidence but retained as background on mechanism.
Criteria may be refined after a pilot screen. If they change, rescreen earlier records and document the change.
Next steps: Screen Titles and Abstracts With Written Criteria
Screen twenty records independently or with a partner, compare disagreements and improve ambiguous criteria before scaling.
Trace Study Families Before Counting Evidence
Several articles can report one trial, cohort or dataset. A map may display each paper as a separate node even though the evidence is not independent.
Compare author teams, sample size, recruitment dates, setting, registration numbers and baseline characteristics. Link companion methods, outcome and follow-up papers.
A practical example: Trace Study Families Before Counting Evidence
Three nodes describe immediate, six-month and qualitative outcomes from the same classroom programme. The review treats them as one study family with several reports, not three replications.
Similar samples can be difficult to identify when author lists or titles change. Uncertainty should remain visible.
Next steps: Trace Study Families Before Counting Evidence
Add a study-family field to the evidence table and count independent samples separately from publications.
Search Deliberately for Counterevidence
Recommendation systems often expand around what the seeds already represent. A literature review becomes stronger when the reader actively looks for null results, failed replications, alternative theories and critical reviews.
Use contrasting seeds, negative-outcome terms and references cited in disagreement. Inspect what influential reviews excluded.
A practical example: Search Deliberately for Counterevidence
A map built from positive classroom studies largely returns similar work. Adding a null-result seed reveals a separate methodological conversation about measurement timing.
Absence of counterevidence in the map does not prove consensus. Publication and citation practices can suppress visibility.
Next steps: Search Deliberately for Counterevidence
For every provisional conclusion, write one search designed to challenge it and record what was found.
Calibrate the Size of a Literature Review
A map can expand indefinitely. A stopping rule protects time and prevents a student from equating endless collection with rigour.
Stop when the assignment question is adequately represented across key concepts, methods and viewpoints; recent alert rounds produce no material change; or the protocol’s date and source limits are reached.
A practical example: Calibrate the Size of a Literature Review
A school project stops after twelve fully read sources covering review, direct classroom studies, mechanism and limitation. A dissertation requires a broader protocol and formal screening record.
Stopping is a judgement, not proof that nothing remains. State the boundary honestly.
Next steps: Calibrate the Size of a Literature Review
Set the rule before the final search and ask whether new sources change the synthesis or merely repeat it.
Turn a Map Into a Reading Sequence
Students often open papers randomly. A deliberate sequence can reduce confusion: overview, vocabulary, methods, direct evidence, counterevidence and recent updates.
Read one strong review to map terms, then representative empirical papers from different clusters. Return to foundational works only when the later literature identifies their role.
A practical example: Turn a Map Into a Reading Sequence
A learner first reads a review, then one classroom trial, one laboratory mechanism paper and one critical synthesis. The map now supports a coherent learning journey rather than tab accumulation.
Reviews can transmit the biases of their included literature. They orient but do not replace primary-source checks.
Next steps: Turn a Map Into a Reading Sequence
Write a five-source reading queue and a reason for the order. Change it when a paper proves less relevant than expected.
Explain the Map in an Oral Presentation
A map can be useful in a presentation if the speaker explains what it represents and avoids decorative authority.
State the seeds, relationship, date, filters and limitation. Use the visual to show search strategy or field structure, then move to source-level evidence.
A practical example: Explain the Map in an Oral Presentation
The student says, “This view groups records through shared citation patterns; the highlighted cluster guided our reading, but our conclusion comes from the four studies summarised next.”
A screenshot without legend, method or accessible description can confuse viewers and exclude those who cannot see colour differences.
Next steps: Explain the Map in an Oral Presentation
Provide a text description, define the relationship and cite the underlying papers beside every substantive claim.
Did You Know? Newer Papers Can Look Quiet
Citation networks accumulate over time. A recent article may have few visible links even when it is highly relevant and carefully designed.
Balance network discovery with date filters, Similar Text, alerts and direct database searches. Read relevance and method before interpreting low connectivity.
A practical example: Did You Know? Newer Papers Can Look Quiet
A 2026 classroom study appears at the edge of the map because few later papers have cited it. Its population and outcome match the question better than several large older nodes.
Low citation counts can also reflect niche scope, language, access or indexing delay. None of these automatically establish quality.
Next steps: Did You Know? Newer Papers Can Look Quiet
Create a recent-work search independent of node prominence and compare its results with the central clusters.
Frequently Asked Questions
Does a Litmaps map show the best papers?
No. It shows records connected under selected inputs, algorithms and views. Quality, relevance and evidential strength require reading and evaluation.
What is a good seed article?
A verified source chosen for a stated reason: broad review, close empirical match, foundational method or contrasting approach. More than one seed reduces dependence on a single lens.
Are nearby nodes direct citations?
Not necessarily. Proximity depends on the selected view and algorithm. Check the records and direct citation relationships before describing the connection.
Can Litmaps replace a library database?
No. It can complement database searching and citation chasing, but coverage, document types, languages and metadata have limits.
Are Monitor alerts always newly published research?
No. A record can be newly added to the database while the paper itself is older. Verify publication and version dates.
Can a blank map area prove a research gap?
No. Test alternative seeds, terminology, languages, databases and local sources before making a narrow, evidence-based gap claim.
Should students cite Litmaps?
Cite the underlying papers for substantive claims. Describe the mapping method when it materially shaped discovery, following the assignment’s method and citation rules.
What should a parent ask?
Ask what the map relationship means, which paper was actually read, what claim it supports and what the search may have missed.
Why do two Litmaps searches from the same seeds differ?
Database updates, filters, algorithm settings, account features and changed input sets can alter results. Record the date, seeds, settings and screening decisions so a later reader can understand the workflow even when the live map changes.
How should a student describe a map in a presentation?
State what connects the nodes, which seeds and settings were used, and what the map cannot establish. Use it to explain discovery or field structure, then present source-level evidence for conclusions.
Useful Next Reading
- Litmaps: Create a Litmap — verify the current seed and exploration workflow.
- Litmaps search algorithms — compare citation, author and text-based discovery.
- Reading Connected Papers similarity graphs — compare a different mapping mechanism without creating a second broad literature-map hub.
- Critical reading, media and information literacy — follow a claim from interface to source.
Closing thought: A good literature map expands curiosity without replacing judgement. English supplies the labels, limits and relationship verbs that turn an attractive network into a transparent research process.