HOW SINGAPORE EDUCATION WORKS · University Literature Search and Keeping a Search Trail
Find your next learning step
Choose the question closest to your situation, or read the full guide in order.
Full chapter index · Practice and checking · Sources and next reading · How Singapore Education Works
For a university literature search, turn your question into two or three main concepts, list alternative terms for each concept, and try a search you can explain. Read a few relevant results, refine one part of the search, and record what changed. Finding more items is useful only when you can judge how they help answer the question.
A search trail is a short record of where you searched, the exact query, the date, the filters and your next decision. It helps you resume the work and explain your choices. It does not turn an ordinary assignment into a systematic review or prove that every relevant publication has been found.
This guide is for students beginning a library-based inquiry. Follow your module brief and your own institution's library guidance. The small catalogue below is fictional; its results are calculated from the supplied tags, not reported from a live database.
The choices at a glance
| What you notice | Possible issue | Useful next move |
|---|---|---|
| A search returns too much | The concepts may be too broad | Inspect terms, phrases and search fields |
| A search returns too little | Restrictions or terminology may hide material | Check alternatives and filters |
| Matching words give irrelevant items | Words occur in an unhelpful field or context | Inspect the actual record |
| The search cannot be resumed | Retrieval decisions were not recorded | Keep the exact query and its conditions |
Full chapter index
Define the evidence and choose search terms
Inspect a search case and read the results
Repair retrieval and practise a search log
Resume, transfer and answer questions
CHAPTER 1 OF 8 · Define the evidence and choose search terms
1. Begin with the question and the kind of evidence needed
Suppose your working question concerns how shade relates to outdoor thermal comfort on university campuses. You need sources that help define the concepts, describe relevant methods and report findings you can evaluate.
Before searching, inspect the assignment. Does it ask for a comparison, an argument, a review of a field or a design proposal? What counts as appropriate evidence? A useful policy document may play a different role from a measurement study.
Write a provisional inclusion rule in plain language. For example: “I am looking for research about outdoor thermal comfort in campus settings, including relevant shade or shelter conditions.” This is a working boundary, not a universal research standard.
If you cannot say what evidence could answer your question, pause and use the guide to narrowing a university research question. Better search syntax cannot compensate for an inquiry whose purpose remains unclear.
You may revise the question after an initial search. Record the reason so the change remains visible.
CHAPTER 2 OF 8 · Define the evidence and choose search terms
2. Separate concepts from alternative wording
For the shade example, possible concept groups include the setting, the environmental feature and the human experience. Draft a small vocabulary map before combining everything into a long sentence.
Possible setting terms include campus and university. Feature terms might include shade and shelter, although these are not identical concepts. For experience, thermal comfort is more precise than the everyday word “comfortable.”
Treat alternatives as candidates to test. A search result can teach you the terminology used by a discipline. A relevant paper may use a phrase you did not anticipate, while a familiar word may retrieve a different topic.
In many library search systems, OR connects alternatives and AND combines concepts. Quotation marks can request a phrase. Check the particular system's help because indexing, matching, phrase handling and operator rules differ.
The NUSearch+ guide documents those tools for NUS's discovery service. That is an institutional example, not a guarantee that the same query will behave identically in another university's catalogue.
Start simple enough that you can tell which change affected the results.
CHAPTER 3 OF 8 · Inspect a search case and read the results
3. A worked case: search a six-record practice catalogue
Assume this fictional practice catalogue searches only the listed tags. Matching is exact, and the operators have ordinary set meanings: AND requires both groups; OR allows either term. No stemming, ranking, synonyms or full-text matching applies.
| Record | Supplied tags |
|---|---|
| A | campus, shade, comfort |
| B | street, tree, temperature |
| C | campus, shelter, comfort |
| D | campus, shade, energy |
| E | park, shade, comfort |
| F | campus, shade, temperature |
The query shade AND campus retrieves A, D and F: three records. B lacks both tags, C lacks shade, and E lacks campus.
Now use (shade OR shelter) AND campus. It retrieves A, C, D and F: four records. C is added because shelter satisfies the first group. You broadened one concept group while keeping the campus condition.
Next use (shade OR shelter) AND campus AND comfort. Only A and C remain: two records. The additional AND condition narrows the set to records that also carry comfort.
The counts are not research findings. They show how this deliberately simple search logic works. Record C still needs reading: a shelter discussion might be relevant or might concern a different issue. A matching term is an invitation to inspect a source, not evidence that it answers your question.
CHAPTER 4 OF 8 · Inspect a search case and read the results
4. Read results before deciding how to revise
Open a promising record and inspect its title, abstract, publication details and available full text. Ask what it studies, where, using what evidence, and with what limitations. Keep a short relevance note separate from any judgment about quality.
For a source that seems relevant, record a stable identifier or library link and enough bibliographic details to find it again. Check exported citations against the actual item; automated formatting can still contain errors.
If many results concern energy use rather than human comfort, adding or adjusting the experience concept may help. If useful studies disappear, remove that restriction and inspect their vocabulary before tightening again.
Do not add NOT casually to make a large number smaller. Australia's CSIRO Library search guide warns that exclusions can remove relevant references. A paper may mention an unwanted topic while still containing evidence you need.
A relevant reference list can suggest earlier work. Citation tools may point to later work where available. Follow these routes as additional leads and record how you found them, rather than pretending they came from the original query.
CHAPTER 5 OF 8 · Repair retrieval and practise a search log
5. Diagnose too many, too few or misleading results
If there are too many results, check whether the query combines your actual concepts. A broad word such as comfort may retrieve unrelated uses. A phrase, appropriate field search or justified filter may help, depending on the database.
If there are too few, inspect spelling, overlong exact phrases, restrictive dates and missing alternatives. Check that your AND groups do not require a term that relevant authors rarely use.
If apparently irrelevant results appear, inspect where the matching words occur. A campus name in an author's address is different from a campus study setting. Search fields can matter as much as the words.
If a useful item is unavailable, use your institution's permitted access routes or ask a librarian. A record existing in search results does not guarantee that its full text is available through your account.
If you only find papers agreeing with your preferred answer, review the query's framing. Words that presuppose a benefit may hide studies discussing limits or trade-offs. Search for the relationship being investigated, not only the conclusion you hope to reach.
The next task is academic reading and evaluating arguments, once you have sources worth reading closely.
CHAPTER 6 OF 8 · Repair retrieval and practise a search log
6. Practise a revision and keep a usable log
For guided practice, use the six-record catalogue and search shade AND comfort. Identify the matching records before reading the answer.
Checking answer: A and E match, giving two records. D and F include shade but not comfort. Adding AND campus leaves A alone. The change affects the permitted setting, not the meaning of the shade tag.
For independent practice, compare campus AND (shade OR shelter) with campus AND shade AND shelter. Explain the difference in words and list the results.
Checking answer: the first retrieves A, C, D and F. The second retrieves none because no record has all three tags. A zero result here follows from the supplied catalogue. It says nothing about whether real campus shelters exist or whether research has been conducted.
For a real assignment, record a compact log entry: search date; database or discovery service; exact query; fields and filters; observed result count; promising records; and the reason for the next revision. Record actual counts only after running the search.
Keep a separate source note about what an item contributes. The search log explains retrieval decisions; the reading note explains evidence and interpretation.
CHAPTER 7 OF 8 · Resume, transfer and answer questions
7. Return to the search and adapt across disciplines
At your next study session, use the log to reconstruct the last useful query before opening a new search. Can you explain why you added each concept and filter? If the reason is missing, revisit the decision instead of copying it automatically.
For changed-context practice, imagine researching wayfinding in museums rather than thermal comfort on campuses. Keep the method of identifying concepts and alternatives, but build new vocabulary. Carrying over shade-related terms would preserve the old topic rather than transfer the searching skill.
The Singapore and Australian library guides show a useful global connection: researchers share search principles while working with different collections and interfaces. Record the platform as well as the query, because a search statement alone may not reproduce the same result elsewhere.
Different disciplines also organise knowledge differently. Books, archives, conference papers, datasets and journal articles can serve different purposes. Ask your subject librarian which sources fit the inquiry.
For a literature review with a formal protocol, obtain the required methods guidance. The modest log in this article supports accountable study; it does not establish the completeness expected of a systematic review.
CHAPTER 8 OF 8 · Resume, transfer and answer questions
8. Common questions about university literature searching
Should I type the entire assignment question into the search box?
That can be an exploratory start in some systems. For a search you need to refine and explain, separating the concepts often makes the next decision clearer. Compare the results rather than assuming one style always works best.
Is the first result the strongest evidence?
Search ranking is not an assessment of suitability for your particular argument. Inspect the item, method and relevance. A highly ranked source may provide background rather than the evidence your claim needs.
How many references are enough?
Use the assignment requirements and the needs of the argument. A fixed number cannot show whether relevant perspectives, methods or limitations have been considered. Ask your lecturer when the brief is unclear.
Does a zero-result search prove there is a research gap?
No. The database, terminology, fields and filters may explain the result. Try justified alternatives, consult library help and inspect relevant source trails before making a broad absence claim.
Can I use an AI-suggested title?
Treat it as an unverified lead. Locate the actual item, confirm its bibliographic details and read what it says. Follow your institution's rules about AI assistance and authorship.
Sources and next reading
Sources checked on 22 October 2026: NUS Libraries, NUSearch+ and CSIRO Library, Creating effective search strategies. The six-record catalogue, calculations and practice tasks are original fictional examples.
Continue with narrowing a research question and academic reading, or return to How Singapore Education Works.
Learning at other stages
For early-years learning, explore fair comparisons and equal measuring units. Comparing Lengths and Measuring in Kindergarten.
For school mathematics, distinguish examples, counterexamples and a general argument. Examples, Counterexamples and Mathematical Proof.
For work, volunteering and later-life learning, make a task understandable to another person. Writing Clear Instructions for Adult Learners.