How do you build a research question, keyword map and library search strategy? Start with the question in ordinary language, identify the concepts that carry its meaning, build synonyms around those concepts, and search iteratively. Each search should teach you something: a better term, a useful author, a subject heading, a missing population or a reason to narrow the question.
This is the difference between typing and researching. Typing asks the database to understand whatever sentence you happen to write. Researching studies how the collection is organised, learns the vocabulary used by experts, and changes the next search deliberately. A search strategy is therefore not one magical string. It is a documented sequence of increasingly informed decisions.
Careful organisation keeps the question and evidence clean. Evidence-based observation notices recurring words, names and patterns. Cross-disciplinary synthesis connects those clues into a topical map. The result is a happy, practical research habit: you know why you searched, what changed, and when it is time to stop searching and begin understanding.
Research a Topic with Databases, Keywords and Credible Sources
A Topic Is Not Yet a Question
Climate change, artificial intelligence and Singapore history are territories. A research question gives the territory direction and boundaries. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Write the Question in Ordinary Language First
State what you genuinely want to know before translating it into database syntax. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Identify the Core Concepts
Underline the ideas that must be present in a useful source and remove filler words. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Separate Concepts from Limiters
Age group, country, date and source type can narrow a search, but adding every limiter immediately can hide useful vocabulary. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Build Synonym Families
For each concept, list everyday words, academic terms, spelling variants, acronyms and related phrases. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Search Broadly to Learn Vocabulary
The first search can be reconnaissance. Notice recurring subject terms, authors, journals and phrases. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Search Narrowly to Test a Question
Once vocabulary improves, combine concepts and appropriate limiters to find material closer to the need. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Use Exact Phrases Selectively
Quotation marks can help with distinctive phrases but can also exclude relevant wording. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Understand AND, OR and NOT
Where supported, AND combines concepts, OR expands synonyms and NOT excludes—but careless exclusion can remove useful evidence. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Use Filters After Understanding Results
Dates, languages, source types and peer-review filters are useful when they solve a real problem. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Read Subject Headings and Keywords
Good records teach you vocabulary used by the collection and research community. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Keep a Search Log
Record useful search strings, databases, filters and what changed. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Change One Thing at a Time
Adjust one concept, synonym or limiter so you can learn which change improved the results. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Use Anchor Sources
One strong overview can reveal vocabulary, authors and references that make the next search smarter. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Search Forward and Backward
References reveal earlier work while cited-by tools reveal later work. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Test for Confirmation Bias
Search for limitations, criticism, alternatives and contrary evidence rather than only support. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Know When the Question Is Too Broad
If results scatter in many directions, narrow the population, mechanism, place, period or outcome. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Know When the Question Is Too Narrow
If reasonable synonyms produce almost nothing, remove unnecessary limits or reconsider assumptions. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Stop Searching and Start Reading
When new results mostly repeat known sources and main claims have evidence, shift to understanding. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Turn Search Notes into an Outline
Group sources by the sub-question they answer rather than the order you found them. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
A Singapore Library Search Practice
Use OneSearch and eResources to test how the same concepts behave across books, newspapers, eJournals and other collections. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
Common Mistakes and Repairs
Avoid one giant sentence, endless synonym swapping without notes, premature filters and collecting results without reading. Write a small example from your own topic before opening the database. This forces the decision to become visible and gives you something concrete to revise.
During the search, look at more than result counts. Read titles, subject terms, abstracts and metadata. Ask which word is doing useful work and which word is merely adding noise. Record promising vocabulary in a keyword map so that the next query is based on evidence from the collection.
For example, a student researching artificial intelligence and writing may discover that researchers use terms such as generative AI, large language models, writing assistance, academic integrity or automated feedback. Those terms divide one vague topic into several different research questions. The vocabulary is helping the student see the field.
Change the search deliberately. Add a population only when it matters. Use a date limit because the technology changed, not because a filter is available. Remove a quotation mark if it blocks useful variants. Search a broader term when the question is too narrow. Each adjustment should have a reason.
Finish with a short note: search used, tool or database, what you learned, which sources are worth reading and the next action. This is enough to prevent circular searching and makes it possible to explain your research method later.
A Reusable Keyword Map
- Concept 1: the main phenomenon or subject.
- Concept 2: the outcome, relationship or second idea.
- Population: who or what is being studied.
- Place: include only when geography matters.
- Period: include only when time matters.
- Synonyms: everyday and academic alternatives.
- Related terms: useful neighbouring concepts.
- Exclusions: use cautiously and only for a clear reason.
Build the map on paper or in a simple note. It does not need to become a complicated diagram. Its job is to keep alternative language visible so that a failed search becomes a prompt to test another route rather than a reason to conclude that no evidence exists.
A Worked Example: From Social Media to a Researchable Question
Begin with the topic social media. Ask what about it matters. Suppose the interest is sleep among teenagers. A first question might be: How is late-night social media use associated with sleep quality among adolescents? The concepts become social media use, sleep quality and adolescents. Late-night use may become an additional limiter.
Now build synonyms: social networking, digital media, platform use; sleep duration, sleep disturbance, bedtime; teenagers, youth, secondary-school students. Search several combinations and notice which terminology appears in strong sources. The search is teaching you how researchers describe the phenomenon.
If the results are still enormous, narrow intelligently: perhaps a country, a study type or a more precise outcome. If the results disappear, remove a limiter. The research question and search strategy develop together.
A Worked Example: Singapore Local History
Start with a place rather than a theme: Clementi, Bras Basah or Punggol. Decide what change interests you—transport, schools, housing, commerce, shoreline or community life. Then build historical place-name variants and a date window.
Search books, newspapers, maps and records differently. A newspaper archive may respond to a street name and date; a map collection may require a place or geographic feature; a catalogue may surface a book under a broader subject heading. One research question can therefore need several search grammars.
NLB OneSearch exposes books, newspapers, magazines and articles, records and papers, images, websites and eJournals through one discovery environment. Use it to compare how the same concept behaves across source types.
Frequently Asked Questions
How many keywords should I use? Enough to express the core concepts without making the query brittle. Start with a small combination and add limits deliberately.
Should I search my whole question in quotation marks? Usually not. Exact-phrase searching can be useful for distinctive wording, titles and names, but a full quoted question may exclude relevant scholarly language.
What does OR do? In systems that support Boolean operators, OR usually expands a concept across alternatives, such as adolescent OR teenager. Check the database’s own search help.
When should I stop? When new searches mostly repeat known sources, the important sub-questions have appropriate evidence and further discovery is unlikely to change the answer materially.
Why keep a search log? It prevents repeated dead ends and shows how your vocabulary and decisions evolved.
Helpful Reading
- Find Academic Journals, Peer-Reviewed Articles and Research Papers
- Use Google Scholar with Library Access
- Evaluate Sources for Credibility, Authority, Bias and Evidence
- NLB OneSearch
How to Use a Library: Search Until the Question Becomes Clearer
The best search strategy does not merely produce results. It improves the question. You learn the vocabulary, see the boundaries of the evidence, discover disagreement and understand which source types are capable of answering each part.
Write the question. Map the concepts. Search broadly enough to learn. Narrow with a reason. Keep a log. Follow strong sources. Search for disagreement. Then stop searching and read. That is how a library search becomes a repeatable method rather than a lucky query.
