Aug 10, 2026

How to Use Boolean Search for Research Papers: AND, OR, NOT, and Search Strings

By JournalLabs Research Team

Introduction

Finding research papers requires more than entering a topic into a search box.

Researchers usually begin with a question written in ordinary language, but academic databases retrieve papers through the terms, phrases, fields, and operators included in a query. A relevant paper may use terminology that differs from the researcher’s wording, while a broad search may return thousands of records that mention the topic without addressing the actual question.

Boolean search helps researchers control this process.

The operators AND, OR, and NOT define how concepts should be combined. Quotation marks preserve exact phrases, parentheses organize alternatives, and a structured search string makes it easier to balance broad coverage with relevant results.

For example, a researcher interested in remote work and productivity might search:

("remote work" OR telework OR "working from home")

AND

(productivity OR "job performance")

This query groups alternative language for each concept and requires results to address both.

Boolean search is especially useful when researchers need a transparent and repeatable strategy. Strong search strings are usually refined after researchers inspect early results and learn how the field describes the topic.

This guide explains how Boolean operators work, how to build and improve academic search strings, and how AI-assisted literature search can support the process without replacing researcher judgment.

What Is Boolean Search?

Boolean search combines search terms with logical operators.

The three main operators are:

• AND, which requires different concepts
• OR, which allows alternative terms
• NOT, which excludes a specified term

Consider a search about social media and adolescent mental health.

A simple query might be:

social media adolescent mental health

A Boolean version could be:

("social media" OR "social networking" OR "digital media")

AND

(adolescent OR teenager OR youth)

AND

("mental health" OR depression OR anxiety)

The structured version recognizes that papers may use different terminology for the same idea. It also requires records to connect the main areas of interest.

Academic databases support similar Boolean logic, but syntax and field codes may differ. Researchers should preserve the concepts while adapting the query to each platform.

Why Boolean Search Matters for Research Papers

Research questions contain concepts rather than one fixed phrase.

For example:

How does online learning affect academic performance among university students?

The central concepts are:

• Online learning
• Academic performance
• University students

Each concept can appear under alternative terms. Online learning may be described as e-learning or distance education; academic performance may appear as academic achievement or learning outcomes; university students may be called college students or undergraduates.

Searching only the original wording may miss relevant studies, while adding terms without structure may create noise. Boolean logic expands the language within each concept while requiring connections between the main concepts.

It also makes search decisions explainable. A documented query records which concepts were used and how they were combined, making the search easier to repeat or update.

For a formal literature review, researchers may also need to record:

• Which databases were searched
• Which keywords and subject terms were used
• Which filters were applied
• When the search was performed
• How the search strategy changed

A clear Boolean string provides a visible record of the search logic.

How AND, OR, and NOT Work

Use AND to Connect Different Concepts

AND narrows a search by requiring each connected concept.

For example:

"machine learning" AND "hospital readmission"

This retrieves records connected to both topics.

Adding another concept narrows the query further:

"machine learning" AND "hospital readmission" AND adults

AND should connect distinct parts of the question. Too many AND conditions can make a search overly restrictive.

A relevant paper may address the intended population or context without using the exact term in its title, abstract, or keywords.

Use OR to Combine Synonyms

OR broadens a search by allowing any connected term to appear.

For example:

adolescent OR teenager OR youth

OR is useful for synonyms, abbreviations, alternative spellings, and related expressions. Group them with parentheses:

(adolescent OR teenager OR youth)

This shows that the terms represent alternatives within one concept.

OR may also connect:

• Technical and plain-language terms
• Older and newer terminology
• Singular and plural variations
• Regional spelling differences
• Abbreviations and full terms

Use NOT Carefully

NOT excludes records containing a term.

For example:

jaguar NOT car

In research searches, NOT can remove useful papers. A study focused on adults may mention children in the background and be excluded even though its main sample is relevant.

Use NOT only when an unwanted meaning dominates the results and the exclusion has been tested.

Researchers should examine some of the excluded results before keeping a NOT condition. Narrowing the required concepts is often safer than excluding a broad term.

How to Build a Boolean Search String

Step 1: Define a Focused Question

Instead of:

Artificial intelligence in education

use:

How do AI-based feedback tools affect writing performance among university students?

This question contains three central concepts:

1. AI-based feedback tools
2. Writing performance
3. University students

Not every detail must appear in the final query. The goal is to identify the concepts that distinguish relevant papers from unrelated records.

Step 2: Build a Concept Table

List each concept with possible search terms.

A concept table makes missing terms, ambiguous words, and unnecessary restrictions easier to identify.

Terms can come from:

• The research question
• Relevant papers
• Subject headings
• Review articles
• Common abbreviations
• Checked AI suggestions

Step 3: Group Synonyms with OR

Create one group for each concept:

("artificial intelligence" OR AI OR "automated feedback" OR "intelligent tutoring")

(writing OR composition OR "writing quality" OR "writing achievement")

(universit* OR college OR undergraduate* OR "higher education")

Each group represents one idea using several possible expressions.

Step 4: Connect Concept Groups with AND

Combine the groups:

("artificial intelligence" OR AI OR "automated feedback" OR "intelligent tutoring")

AND

(writing OR composition OR "writing quality" OR "writing achievement")

AND

(universit* OR college OR undergraduate* OR "higher education")

The query now allows alternative language within each concept while requiring the results to address all three areas.

Step 5: Test the Search

Do not assume the first query is complete.

Review the results and ask:

• Are known relevant papers retrieved?
• Which terms appear in useful papers?
• Which terms create irrelevant results?
• Is one concept too restrictive?
• Are important synonyms missing?
• Does the database provide useful subject headings?

A search string should be evaluated using actual results, not only by how complete it looks.

Step 6: Record Revisions

Save:

• The complete query
• Database or platform
• Search date
• Filters
• Result count
• Major changes

Record what changed and why so the search can be audited or updated later.

How to Use Quotation Marks, Parentheses, and Truncation

Quotation Marks

Quotation marks usually keep words together:

"systematic review"

They are useful for established phrases, but excessive quotation can miss papers that use a slightly different word order.

For example, searching only:

"artificial intelligence feedback"

may miss papers using:

AI-generated feedback

Quotation marks should therefore be used selectively.

Parentheses

Parentheses group related terms:

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

They clarify which terms are alternatives and which concepts must be connected.

Without parentheses, the database may process mixed AND and OR conditions in an unintended order.

Truncation

Many databases use a symbol, often an asterisk, to retrieve word variations.

For example:

adolescen*

may retrieve:

• Adolescent
• Adolescents
• Adolescence

Truncation can improve coverage, but short roots may retrieve unrelated words. Researchers should check each database’s rules before using it.

Field Searching

Some platforms allow researchers to search within:

• Titles
• Abstracts
• Author names
• Keywords
• Subject headings
• Publication names

A title search may improve precision but reduce coverage. Field codes differ across databases and should be adapted carefully.

How to Refine a Search

When a Search Is Too Broad

Common signs include:

• Thousands of weakly related results
• Unrelated populations or disciplines
• A dominant alternative meaning
• Papers that mention the topic only briefly

Possible changes include:

• Add an essential concept with AND
• Replace an ambiguous term
• Remove an overly broad OR term
• Use a more specific phrase
• Search a key concept in the title or abstract
• Apply a justified date, subject, or study-type filter

Do not add many filters at once. Identify which part of the query is creating the noise.

When a Search Is Too Narrow

Common signs include:

• Very few results
• Known relevant papers are missing
• Results are limited to one terminology set
• The query requires details that papers may not mention in searchable fields

Possible changes include:

• Add synonyms with OR
• Remove the least essential AND concept
• Use fewer exact phrases
• Broaden the population or outcome term
• Use truncation
• Search broader fields
• Follow references and citations from a relevant seed paper

The objective is not the smallest result set. It is a manageable set that still represents the relevant literature.

Boolean Search vs AI-Assisted Search

Boolean search provides explicit control over the relationship between terms.

AI and natural-language search allow researchers to describe a question directly:

How do remote working arrangements affect employee productivity and well-being?

Semantic systems can retrieve papers related by meaning, even when they do not use the exact query wording.

The approaches are complementary.

Boolean search is useful for:

• Transparent query logic
• Reproducible searches
• Controlled synonym expansion
• Database-specific searching
• Formal search documentation

AI-assisted search is useful for:

• Exploring unfamiliar topics
• Identifying alternative terminology
• Discovering semantically related papers
• Refining a broad research direction
• Finding useful seed papers

A strong workflow may begin with natural-language exploration, use relevant papers to learn the field’s vocabulary, and then build a Boolean strategy for a more systematic search.

Researchers should still screen the results and verify each selected paper against an original or reliable publication record.

Traditional vs AI-Assisted Search-String Development

AI can accelerate query development, but every suggested term and operator should be reviewed before formal use.

Common Boolean Search Mistakes

Using AND Between Synonyms

This query is too restrictive:

adolescent AND teenager AND youth

It requires all three terms.

Use:

adolescent OR teenager OR youth

when they represent alternatives.

Using OR Between Different Concepts

This query is too broad:

remote work OR productivity

It retrieves papers about either topic independently.

Use AND when both concepts are required.

Adding NOT Too Early

An exclusion may remove papers that mention the unwanted term only as background.

Test the effect before keeping NOT in the final query.

Forgetting Parentheses

Mixed AND and OR queries may be interpreted in an unintended order.

Group synonyms before connecting different concepts.

Quoting Every Phrase

Excessive quotation marks can miss useful wording variations.

Use exact phrases selectively.

Building a Long Query Before Testing

A complex query may fail because one narrow term blocks relevant records.

Start with essential concepts and refine the query iteratively.

Reusing the Same Syntax Everywhere

Databases differ in phrase handling, truncation, subject headings, field searching, and proximity operators.

Adapt the syntax without changing the conceptual structure.

Treating Matches as Final Evidence

A matching paper still requires relevance, design, source, and inclusion review.

Boolean retrieval is the beginning of paper selection, not the final inclusion decision.

How JournalLabs Helps Researchers Search Academic Literature

JournalLabs’ AI Literature Search helps researchers move from an ordinary-language question to a focused and traceable set of academic papers.

Researchers can use JournalLabs to:

• Search with natural-language questions
• Identify central concepts
• Explore alternative terminology
• Refine a research direction
• Review relevance-ranked publications
• Discover connected papers
• Save useful studies
• Return to original source records

For example, a researcher studying remote work and software-team productivity may discover terms such as telework, distributed work, virtual teams, task performance, and employee output. These terms can then support a more precise Boolean strategy.

For the complete discovery workflow, see How to Search Academic Literature with AI: Keywords, Filters, and Source Verification.

JournalLabs supports query development and paper discovery, but the researcher remains responsible for checking suggested terms, adapting database syntax, verifying sources, and selecting the final evidence set.

Frequently Asked Questions

What Are the Main Boolean Operators?

The main operators are AND, OR, and NOT. AND connects required concepts, OR combines alternatives, and NOT excludes a specified term.

Should Synonyms Use AND or OR?

Synonyms should usually be connected with OR so papers using different terminology can appear.

Why Are Parentheses Important?

Parentheses group terms within the same concept and preserve the intended relationship between AND and OR.

Should Researchers Use NOT?

Only cautiously. NOT can reduce irrelevant results but may also remove useful papers that mention the excluded concept.

Are Boolean Searches the Same in Every Database?

No. The logical structure is similar, but field codes, truncation symbols, phrase rules, subject headings, and proximity operators may differ.

Is Boolean Search Better Than AI Search?

Neither is always better. Boolean search offers explicit and reproducible control, while AI search supports exploration and terminology discovery. Combining them is often stronger.

How Can Researchers Evaluate a Search String?

A useful search string retrieves known relevant papers, covers important terminology, and produces manageable noise. It usually needs several rounds of testing.

Conclusion

Boolean search turns a research question into a structured and explainable literature-search strategy.

AND connects different concepts, OR combines synonyms, and NOT excludes unwanted meanings when used carefully. Quotation marks, parentheses, truncation, and field searching provide additional control.

The quality of a query depends less on its length than on its logic. Researchers need to identify the essential concepts, learn the terminology used in the field, test early results, and revise the query when useful papers are missed or irrelevant records dominate.

Boolean search also works best as part of a broader discovery process. AI-assisted search can support exploration and terminology development, while Boolean logic supports precision, reproducibility, and documentation.

The strongest workflow is:

Define the question, organize the concepts, build the query, test the results, refine the terms, and document the decisions.

Search Academic Literature with JournalLabs

Turn research questions into focused searches, discover relevant papers, refine your terminology, and trace every result back to its original source with JournalLabs.

Start searching for research papers with JournalLabs.