03 · What You Need to Know
How Proximity Searching Controls the Relationship Between Search Terms
AND Requires Both Terms but Does Not Necessarily Keep Them Together
Suppose you search:
This is entirely consistent with what the Boolean AND operator actually does. It intersects two search sets. It does not ask whether the terms form a phrase, describe one another, or occur in the same sentence.
For distinct concepts, that may be exactly what you want. If the question concerns diabetes among adolescents, the words representing diabetes and adolescents do not necessarily need to appear next to each other.
For some concepts, however, textual closeness carries useful information.
Proximity Searching Adds a Distance Requirement
A proximity operator tells the search system that the terms must occur within a specified number of words of one another.
A conceptual search might look like:
By imposing a distance condition, proximity searching can remove records in which both words appear but are textually too far apart to represent the relationship you intended.
Proximity Is a Middle Ground Between AND and an Exact Phrase
Consider three ways of searching the same pair of words:
| Approach |
What It Requires |
Typical Trade-off |
| teacher AND burnout |
Both terms satisfy the search conditions but may be relatively far apart |
More flexible, but potentially more irrelevant co-occurrence |
| Proximity search |
The terms occur within a defined distance |
Allows wording variation while requiring a closer textual relationship |
| "teacher burnout" |
The words behave as a phrase according to the database's phrase-search rules |
Potentially precise, but may miss legitimate variations in wording |
Cochrane's current Technical Supplement describes proximity operators as a way to identify terms that occur near one another without necessarily requiring direct adjacency. It notes that proximity searching can be more sensitive than exact adjacency or phrase searching while providing greater control than simply connecting terms with AND.
Why Can an Exact Phrase Be Too Restrictive?
Suppose you search:
"teacher burnout"
That may retrieve records containing exactly that expression, depending on the database. But researchers could write:
-
burnout among teachers;
-
teacher occupational burnout;
-
burnout experienced by teachers;
-
teachers reporting high levels of burnout.
The conceptual relationship is present in all four, but the exact phrase is not.
Proximity searching can allow some linguistic flexibility while still requiring the relevant words to remain close. This is why it complements rather than replaces phrase searching.
Not Every Pair of Concepts Should Be Forced Into Proximity
Proximity is useful only when closeness between the terms is a reasonable signal of relevance.
Suppose your research question concerns the relationship between social media use and depression among university students. The terms representing social media and depression could legitimately appear in different sentences of an abstract. Requiring them to occur within three words of each other might remove highly relevant studies simply because the authors describe the exposure and outcome separately.
In that situation, AND between concept blocks may be more appropriate.
Use AND when
Different concepts need to be represented in the record, but their textual distance is not itself important.
Consider proximity when
The terms form or describe a relationship that is more likely to be meaningful when the words occur near one another.
Proximity Operators Do Not Have Universal Names
This is one of the most important practical complications.
There is no universal proximity operator that can be copied into every research database. Common operator families include NEAR, ADJ, N, W, NEXT, and platform-specific forms.
Even operators with similar names may not count distance or handle word order identically.
EBSCO, for example, currently documents two proximity operators. N finds terms within a specified number of words in either order, while W requires the terms to remain in the order entered. Its example tax N5 reform can retrieve wording with the concepts in either order, whereas a W search retains the specified order.
ProQuest documentation uses forms such as NEAR/n, while its platform can also support ordered proximity through PRE/n in relevant interfaces. The Cochrane Library uses its own proximity syntax. These differences are exactly why the same search syntax does not work in every research database.
Word Order Can Matter
Some proximity operators allow either order:
teacher... burnout
or:
burnout... teacher
Other operators require the terms to occur in the order you typed them.
Whether order should matter depends on the expression. For many conceptual relationships, either order is useful. In other cases, ordered proximity can improve precision.
| Proximity Type |
Possible Match |
What It Controls |
| Unordered proximity |
teacher... burnout OR burnout... teacher |
Distance, regardless of which term comes first |
| Ordered proximity |
teacher... burnout |
Distance plus the order of the terms |
| Direct adjacency |
teacher burnout |
Terms occur next to one another, sometimes with order specified |
Do not infer order behavior from an operator's name. Check the database documentation.
The Number in a Proximity Operator Is Not Universally Comparable
A search written with a distance of three in one interface may not be exactly equivalent to a distance of three elsewhere.
Platforms can differ in whether the number means the maximum number of intervening words, the total span between terms, or another implementation-specific measure. Punctuation and field boundaries may also affect matching.
Consequently, translating NEAR/3 into another database is not always a matter of replacing NEAR with a different operator and keeping the number unchanged.
Watch Out
Do not assume that N5, NEAR/5, ADJ5, W5, and other proximity expressions are equivalent. Verify how each platform counts words, whether order matters, and which fields support the operator before translating the search.
How Close Should the Terms Be?
There is no universally correct proximity distance.
A narrow distance may improve precision but miss legitimate wording. A wider distance can improve sensitivity while gradually approaching the looseness of AND.
Suppose you are searching for research on teacher burnout. A very narrow distance may capture teacher burnout and burnout among teachers but miss a sentence such as burnout was especially prevalent among secondary school teachers. A much wider distance may retrieve records where the words occur in the same abstract but describe separate issues.
The appropriate distance should therefore be tested against actual language in relevant records.
Use Relevant Articles to Choose the Distance
Known relevant records can show you how authors express the concept.
Inspect titles and abstracts. How many words commonly separate the terms? Does the order vary? Are modifiers frequently inserted? Does one expression routinely occur in a phrase while another appears in a different grammatical structure?
Then test several plausible proximity distances. Compare the relevant records retrieved and the irrelevant material introduced.
This is more defensible than selecting a distance because five words sounds suitably scholarly. Search operators, regrettably, are unimpressed by round numbers.
Proximity Can Reduce False Relationships in Long Text Fields
The value of proximity becomes especially apparent when the searchable field contains substantial text.
EBSCO's current search guidance notes that proximity operators can improve relevance particularly when searching full-text books, long journal articles, or newspaper articles. The longer the searched text, the easier it is for two unrelated terms to occur somewhere in the same document merely by chance.
For example, a full-text article might mention teacher education in its literature review and burnout among nurses several pages later. A broad AND search across full text could satisfy both terms even though the article is not about teacher burnout.
A proximity requirement can reduce this kind of accidental co-occurrence.
Proximity Does Not Guarantee a Meaningful Relationship
Two words occurring near each other are more likely to be related than two words appearing far apart, but proximity is still a textual heuristic.
Consider:
Teacher burnout was not examined in this study.
The words are close. The article may still be irrelevant to a search for empirical studies of teacher burnout.
Likewise, the terms may appear together because the authors are distinguishing two concepts rather than studying their relationship.
Proximity improves control over textual co-occurrence. It does not understand the scientific meaning of the sentence.
Proximity Searching Can Work With Alternative Terms
A concept may have several possible expressions on either side of a relationship.
Suppose you want wording in which teacher-related terms occur near burnout-related terms. Conceptually, you might need:
This can become syntactically complex. Build and test smaller components before combining them into a large expression.
Truncation and Proximity May Not Always Work Together
Another reason to verify database documentation is that platforms differ in which search functions can be combined.
PubMed, for example, currently supports proximity searching in the Title and Title/Abstract fields using syntax of the form "search terms"[field:~N]. Its proximity function has restrictions, including limitations on combining proximity searching with wildcard truncation.
If you need both lexical variation and proximity, you may have to write multiple proximity expressions explicitly rather than relying on one truncated form.
PubMed Now Supports Proximity Searching
Older tutorials sometimes state that PubMed has no proximity operator. That information is outdated.
PubMed added proximity searching in 2022. Its current help documentation allows proximity searches in the Title and Title/Abstract fields. A search is structured by placing two or more terms inside quotation marks and adding a field tag with a proximity distance.
A conceptual example is:
This is a useful reminder to check current documentation rather than relying on a search guide written several years ago. Database interfaces have an inconvenient habit of evolving after methods sections have been immortalized in PDFs.
Proximity Can Be More Sensitive Than Phrase Searching
Cochrane's Technical Supplement specifically recognizes the usefulness of proximity operators in systematic searching. A direct phrase requires a tight formulation, while proximity can permit intervening words and, depending on the operator, different word orders.
For a concept expressed in several grammatical forms, proximity can therefore recover relevant records that an exact phrase would miss.
Consider:
-
burnout among teachers;
-
teacher burnout;
-
burnout in secondary school teachers.
A suitably designed proximity search may capture all three without requiring every phrase to be entered separately.
Proximity Can Be More Precise Than AND
At the other end of the spectrum, AND may allow the terms to occur anywhere within the searched fields.
By requiring textual closeness, proximity can remove records where the terms coexist without representing the relationship you intended.
This is why proximity is often described as occupying a useful position between phrase searching and Boolean AND:
Exact phrase High control over adjacency, but potentially too rigid.
Proximity Allows controlled variation while requiring terms to remain near each other.
AND Allows the greatest separation and therefore potentially more unrelated co-occurrence.
That ordering is a useful conceptual model, not a guarantee that result counts will always behave identically across databases. Automatic mapping, fields, indexing, and other platform features can affect retrieval.
Proximity Is Particularly Useful for Variable Multiword Concepts
Some concepts do not have one stable phrase but do have words that tend to appear near each other.
Examples might include relationships such as:
- a profession near an occupational outcome;
- a technology near a particular educational activity;
- a population descriptor near a condition;
- a methodological term near a measurement concept.
In these situations, proximity can represent linguistic flexibility without allowing the terms to drift independently across a long abstract or full-text document.
Do Not Use Proximity to Repair a Poorly Defined Concept
If your search terms are conceptually wrong, proximity will not make them right.
Suppose you have included a term that is much broader than your intended concept. Requiring it to occur near another word may improve precision, but it can also conceal the underlying vocabulary problem.
First make sure the terms actually represent what you intend to search. Proximity should refine a meaningful relationship between terms, not compensate for a concept block that was poorly constructed.
Test Proximity Against Known Relevant Records
A useful proximity search should retain important relevant records while removing at least some of the irrelevant co-occurrences produced by a looser strategy.
Compare:
Version 1 Search the terms with AND.
Version 2 Search them as an exact phrase where appropriate.
Version 3 Test one or more proximity distances.
Compare Inspect known relevant records, records unique to each version, and the kinds of irrelevant results removed.
If the proximity version loses relevant records because authors commonly separate the concepts more widely than expected, increase the distance or return to AND. If it adds little precision, the operator may not be worth the added complexity.
The Best Proximity Distance May Change Across Databases
Even when the conceptual relationship remains identical, the appropriate implementation can differ because databases expose different fields and count proximity differently.
A search developed in one interface should therefore be translated functionally. Ask what the original proximity expression was intended to achieve, then reproduce that behavior as closely as possible in the destination system.
This is part of the larger task of translating a search strategy between databases. The goal is equivalent retrieval logic, not identical punctuation.