Contents
A good search should start with a well-defined question, but disappointing results can still occur when different terminology is used by sources or when a search tool gives a different interpretation to your words than you meant.

Boolean searches allow you to show the relationships that exist between different concepts; you can state which ideas should be grouped together, include alternative terms, and exclude meanings that you do not want.

That doesn’t make a source reliable or ensure full coverage; instead, it provides you with a more thoughtful method of finding information and a search strategy which you can explain, evaluate, and improve.

The guide will take you through the basic operators and then on to platform-specific searching as well as a responsible AI-assisted research workflow.

What Is Boolean Search?

Boolean search combines terms using logical operators:
Operator Purpose Illustrative example
AND Require different concepts together telework AND productivity
OR Include alternatives telework OR telecommuting
NOT Exclude a concept where supported telework NOT recruitment
The simplest rule to remember is:
When you want to give alternative terms within a concept, use OR; when you want to join different concepts, use AND.
Quotes, parentheses, field tags, and filters can all improve a search, but how they work varies by platform.
The query construction is illustrated by the examples in this guide, and they are not presented as having been tested in actual use.

Understanding AND, OR, and NOT

AND Adds Requirements

Imagine that you are looking for information on telework and employee productivity.

telework AND productivity

A system that applies this logic will retrieve matches that include both requirements in the search fields.
While increasing the number of requirements might improve focus, it could also cause useful sources to be overlooked. For example, a paper dealing with employees might use the term ‘workers’ instead of ’employees’.
Before you add another condition joined by AND, consider whether that word actually has to appear.

OR Accounts for Different Wording

Authors don’t stick to the same language every time.

telework OR telecommuting OR "remote work"

It enables other descriptions of the way the arrangement works.

Not all terms related to the subject are synonyms. The term ‘flexible work’ may cover arrangements which have nothing to do with working from home. Only include wider terms when it is appropriate for your question.

NOT Can Remove Useful Evidence

An exclusion such as:

telework NOT pandemic

It is possible to eliminate a significant study which looks at working conditions both during and after the pandemic.
Only include exclusions after you have reviewed the results, and never treat this as a final, automatic step; if you are conducting thorough research, note what you excluded and why.

Caption: AND selects the overlap, OR selects either or both concepts, and NOT excludes the second concept. Supported syntax and search behavior vary by platform.

Grouping, Phrases, and Operator Order

On platforms supporting parentheses, group alternatives before connecting them to another concept:

(telework OR telecommuting) AND productivity

Quotation marks can help search for a particular phrase:

"remote work"

But precise wording can reduce coverage. Searching only that phrase may miss sources using “working remotely.”

Operator order also varies. Bing documents a precedence hierarchy, while PubMed processes Boolean combinations left to right unless parentheses establish groups. Explicit grouping helps make your intent clear. Bing search documentation, PubMed search documentation

Build a Search Around the Research Question

Consider:

What evidence explains how remote working arrangements affect productivity?

1. Separate the Concepts

Identify the working arrangement and the outcome. Do not turn every word in the question into a required search term.

2. Create a Term Map

Concept Candidate terms Caution
Working arrangement remote work, telework, telecommuting Terminology may vary by period and discipline
Outcome productivity, performance Performance is broader than productivity
Population employees, workers, staff Making this mandatory may exclude relevant records

3. Draft a Platform-Appropriate Query

For a database supporting the syntax:

(telework OR telecommuting OR "remote work") AND productivity

Keep the first version manageable.

4. Inspect the Sources

Review more than the result count. Read a sample of records and open useful sources.

Ask:

  • Do these sources address the question?
  • Which terms appear in relevant documents?
  • Are expected sources missing?
  • Is a restriction excluding useful material?

5. Refine One Element at a Time

Changing one condition makes it easier to understand its effect.

Record your decisions. A search strategy should explain how you reached the evidence, not simply preserve a complicated string.

Worked Example: Improving a Research Query

The following example illustrates query design on a database supporting AND, OR, and parentheses. It does not report observed search results.

Start with a Simple Query

"remote work" AND productivity

This connects the working arrangement with the outcome. However, it depends on sources using the phrase “remote work.”

Include Alternative Terminology

("remote work" OR telework OR telecommuting) AND productivity

The revised query includes alternative descriptions without requiring every expression to appear.

Change Reason What to check
Add telework and telecommuting Accommodates different terminology Whether the terms fit the question
Connect alternatives with OR Allows any of the descriptions Whether broader coverage introduces irrelevant material
Group alternatives with parentheses Keeps the concept together Whether the platform supports this syntax
Keep productivity as a separate requirement Maintains the outcome focus Whether useful sources use different outcome terminology

Review Before Broadening Further

You might consider adding performance:

("remote work" OR telework OR telecommuting) AND (productivity OR performance)

But performance is not the same as productivity. It may refer to quality, individual assessments, or organizational outcomes.

Add it only if your scope includes those meanings. More results are not automatically better results.

Document the Actual Test

When you run the query, record the platform, date, fields, filters, and what you observed.

Compare a manageable sample of sources from each version. Explain which relevant sources became easier to find and which irrelevant categories appeared. Do not infer improvement from result counts alone.

Adapt the Search to the Platform

Google Search

Use documented controls rather than assuming Google executes a database query literally.

Examples:

site:ilo.org "remote work"

"remote work" -jobs

"remote work" after:2025-01-01

Google documents phrase searching, site restrictions, exclusions, and date controls. Do not insert spaces after commands: use site:ilo.org, not site: ilo.org.

Check dates on the underlying documents rather than relying solely on search filtering. Google Search Help

Bing

Bing explicitly documents Boolean operators and grouping:

(telework OR telecommuting) AND productivity

Its documentation also states that it uses only the first ten terms to retrieve results. Avoid copying lengthy database strategies into Bing without simplifying them. Microsoft’s advanced search options

Academic Databases

Research databases may provide field searching, subject headings, proximity searches, and truncation.

These are not universal commands. Learn the database’s syntax before using them.

For example, PubMed supports title/abstract tags and lets you inspect Search Details to understand query translation. Restricting fields changes what the search can retrieve. PubMed Help

Recruitment and Monitoring Tools

Check the documentation for the exact product, subscription, and search field.

A platform’s keyword box may behave differently from its title filter. Operator limits and supported features may also differ.

Copyable Search Examples

These examples are starting points, not guarantees of relevant or trustworthy results.

Platform Query Purpose and limitation
Google site:ilo.org "remote work" Focuses discovery on one organization’s site; excludes material elsewhere
Google "remote work" -jobs Excludes jobs; may also exclude relevant sources mentioning employment
Google "remote work" after:2025-01-01 Applies a date restriction; verify dates in the documents
Bing (telework OR telecommuting) AND productivity Connects alternative terms with an outcome; inspect relevance
PubMed (telework[tiab] OR telecommuting[tiab]) AND stress[tiab] Searches title/abstract fields for working arrangements and stress

The PubMed example addresses stress rather than productivity. Changing the outcome changes the research scope—not merely the syntax.

Advanced Techniques: Use Them for a Reason

Field Searching

A title search can reduce incidental matches. But useful material may discuss your concept without naming it in the title.

Compare a field-restricted search with a broader version before relying on it.

Truncation and Wildcards

Some databases allow a symbol to retrieve word endings or spelling variations.

Check the supported symbol, minimum word length, and expansion limits. A short stem can capture many unrelated words.

Do not assume an asterisk works identically in web search and research databases.

Proximity Searching

Proximity searching looks for terms near each other rather than anywhere in a record.

It can help when the relationship between terms matters, but the commands and distance rules vary. Use documented database syntax—not a generic list of interchangeable NEAR, WITHIN, or AROUND commands.

Subject Headings

Controlled vocabulary can complement keywords in databases that support it.

Use both deliberately: keywords capture authors’ wording, while subject headings reflect the database’s indexing system. Neither guarantees complete coverage alone.

Balance Precision and Coverage

A focused search produces fewer irrelevant matches. A broad search may uncover relevant terminology and unexpected evidence.

The right balance depends on the task.

For a quick factual check, a narrow search may be sufficient. For a literature review, broader discovery and careful screening are often necessary.

One useful check is to identify a known relevant source. If your strategy misses it, investigate why. This is a diagnostic check—not proof that the search is comprehensive.

Use AI to Assist, Not Replace, Search Judgment

AI can help draft synonym lists or explain a query. Ask it to separate close synonyms from broader concepts and identify assumptions.

A useful prompt is:

Help develop a search strategy for this question: [question]. Separate the essential concepts, suggest terminology, and explain broader terms that could change the scope. Draft a query for [named platform]. Flag syntax that requires checking against official documentation.

Then review and run the query yourself.

Do not assume Boolean words in a conversational prompt become strict retrieval controls. Unless the tool documents that behavior, they may simply be interpreted as instructions.

A practical combined workflow is:

  1. Define the question and source requirements.
  2. Construct and run platform-specific searches.
  3. Select and read relevant sources.
  4. Use AI to organize or summarize selected material.
  5. Check important statements against the original evidence.
  6. Record unresolved uncertainty.

Summarization and verification are separate tasks.

Before and After: Fixing a Query

Suppose you are researching whether teleworking arrangements relate to productivity.

Before

telework AND telecommuting AND productivity

This requires both descriptions of the arrangement. A useful source using only one term could be excluded.

After

(telework OR telecommuting) AND productivity

This allows either description while retaining the productivity requirement.

The improvement is logical, not a claim about measured search performance. Whether it produces better evidence depends on the platform, collection, and sources retrieved.

Troubleshooting Your Search

Problem Possible explanation Next step
Too few relevant matches Excessive restrictions Remove one condition and compare
Too many unrelated matches Ambiguous or broad alternatives Refine the term map
Missing expected material Terminology or collection gaps Search another source collection
Unexpected behavior Unsupported syntax Check platform documentation
Sources disagree Different dates, definitions, or methods Compare the evidence directly

A technically valid query cannot compensate for a poorly defined question or inaccessible sources.

Keep a Research Trail

Record the platform, search date, exact query, filters, and reasons for changes.

For important sources, preserve identifiers and notes explaining their relevance. Record which claims were verified and which remain uncertain.

Documentation makes the process reviewable. It does not guarantee identical future results from a changing web index.

Research Log Template

Copy this template into a spreadsheet or research notebook. Use one entry for each meaningful search version.

Field What to record
Research question The question the search should help answer
Platform or collection Search engine, database, or tool used
Search date When the query was run
Exact query Copy the query without altering its syntax
Fields and filters Date, language, document type, or field restrictions
Sources reviewed Identifiers or URLs and brief relevance notes
Problems observed Missing sources, ambiguous terms, or irrelevant categories
Query revision What changed and why
Verification notes Claims checked against underlying evidence
Remaining limitations Coverage gaps and unresolved uncertainty

Avoid storing unnecessary personal or sensitive information in the log.

Final Search-Quality Checklist

Before relying on your findings, ask:

  • Question: Does the query reflect what I intended to investigate?
  • Terminology: Have I considered relevant synonyms, abbreviations, and spelling variations?
  • Syntax: Does the platform support the commands I used?
  • Restrictions: Could phrases, exclusions, fields, or filters hide useful sources?
  • Coverage: Should I search another collection, language, or source type?
  • Evidence: Have I read the underlying documents rather than relying on snippets or AI summaries?
  • Interpretation: Have I distinguished what the evidence supports from what remains uncertain?
  • Documentation: Could another researcher understand my decisions?

A completed checklist does not prove that a search is exhaustive. It makes the process more transparent and its limitations easier to identify.

How Can CIRS Certification Improve Your Online Research Skills?

Learning Boolean operators is one part of becoming a capable researcher. The next step is knowing how to select sources, evaluate evidence, document decisions, and use AI responsibly.

AOFIRS’s Certified Internet Research Specialist (CIRS) program provides structured study in online search strategies, research methodologies, source evaluation, AI-assisted research, data analysis, and internet law and ethics.

Practical Example: From Search Query to Research Finding

Suppose you are investigating how remote working arrangements affect productivity. On a platform supporting this syntax, you might begin with:

("remote work" OR telework OR telecommuting) AND productivity

Building the query is only the discovery stage. A structured research approach then involves:

  1. Refining the search: Review terminology and adapt the query to each platform.
  2. Evaluating sources: Examine authorship, publication dates, methods, and supporting evidence.
  3. Comparing findings: Identify differences in populations, definitions, and study conditions.
  4. Using AI carefully: Check summaries against the original documents.
  5. Documenting conclusions: Record the search process, evidence, limitations, and unresolved questions.

This illustrative workflow connects Boolean searching with the broader competencies covered by CIRS. Certification does not replace ongoing practice or guarantee that every finding is accurate.

For further study, explore the AOFIRS Online Search and Research Book alongside the program’s current syllabus and learning materials.

Frequently Asked Questions

Is Boolean search still useful with AI?

Yes. It helps structure searches on systems that support it. AI can assist query development, but source evaluation remains necessary.

Does Boolean search find only reliable information?

No. Matching a query does not establish accuracy, authority, or authenticity.

Can I use the same query everywhere?

No. Adapt operators, grouping, fields, and filters to each platform.

Should every query include NOT?

No. Exclusions are optional and can hide relevant evidence.

What is the difference between Boolean search and advanced search?

Boolean search expresses logical relationships between terms. Advanced search may also include fields, dates, languages, document types, and other controls.

Conclusion

Mastering Boolean search means making thoughtful choices—not collecting the largest number of operators.

Define the question, choose meaningful alternatives, adapt the syntax, and learn from the results. Use AI where it helps, but keep discovery, interpretation, and verification distinct.

The goal is a search process that leads to useful evidence and makes its limitations visible.

Share This Story