How Do You Search for a Research Problem That Uses Different Terminology Across Fields?

How Do You Find Literature When Another Field Calls the Problem Something Else?

You search for a research problem using the terminology you know and find surprisingly little. Then you discover a paper from another discipline describing something remarkably similar with completely different language.

Now the problem changes. You are no longer simply searching for papers. You are trying to discover what other research communities call the phenomenon you care about.

This is particularly important in interdisciplinary research because disciplines develop their own technical vocabularies, theoretical constructs, indexing practices, and preferred ways of framing problems. Searching only the language of your home discipline can therefore make relevant scholarship effectively invisible, even when databases technically contain it.

A good cross-disciplinary search must translate the research problem before it can retrieve the literature adequately.

Search the Concept, Then Translate It Into Each Discipline's Vocabulary

Cross-Disciplinary Searching Is Partly a Translation Problem

Begin with the phenomenon rather than your favorite keyword

Suppose you want to investigate why students continue using generative AI despite knowing that its responses can be inaccurate.

You might begin with terms such as “student reliance on generative AI” or “AI dependence in education.” Those searches are reasonable starting points, but they contain the vocabulary and framing of a particular research context.

Other literatures might discuss related behavior using concepts such as trust in automation, automation bias, reliance on automated advice, algorithm appreciation, cognitive offloading, technology continuance, or advice taking. These concepts are not interchangeable. Some may ultimately prove irrelevant. The point is that searching the underlying behavior can lead you to candidate vocabularies that your initial wording would never retrieve.

Before constructing a search string, write the research problem in ordinary descriptive language. What is actually happening? Who or what is involved? What behavior, process, relationship, mechanism, or outcome are you trying to understand?

This gives you a conceptual anchor that is not owned by any one discipline.

Break the problem into concepts before generating terms

A research question should not usually be pasted wholesale into a database search box. Instead, identify its major searchable concepts.

For example, the question “How does students' trust in generative AI affect their verification of AI-generated academic feedback?” might contain conceptual groups for learners, generative AI, trust or reliance, verification behavior, and feedback.

Not every concept needs to appear in every search. Adding too many concepts with AND can make retrieval excessively narrow. The purpose of decomposition is to identify which parts of the question may have alternative disciplinary expressions.

These are candidate terms, not declarations that the concepts are equivalent. Their usefulness must be checked against the literature.

Let relevant papers teach you the vocabulary

You do not need to know every disciplinary term before beginning. In fact, searching is often iterative precisely because the literature teaches you how the field describes the problem.

When you find a highly relevant paper, inspect more than its title. Look at the authors' keywords, abstract terminology, theoretical constructs, definitions, subject headings, and terminology used in the references it cites.

Then search those terms independently.

This process is less tidy than deciding on five keywords at the beginning and never changing them. It is also much more likely to reveal how another discipline actually thinks about the problem.

Use both keywords and controlled vocabulary

Many bibliographic databases assign standardized subject terms to records. PubMed, for example, uses Medical Subject Headings, or MeSH. Controlled vocabularies help connect records that may use different words for related concepts.

They should usually complement rather than replace free-text keywords.

A new concept may not yet have a sufficiently specific subject heading. Older articles may have been indexed differently. Terminology can change over time. Some databases also use different controlled vocabularies, which means a search strategy cannot always be copied unchanged from one platform to another.

A useful search therefore combines the language authors actually use with the indexing language used by the database.

Search more than one database when the problem crosses fields

A perfectly constructed search cannot retrieve articles that the database does not index.

Database selection should therefore follow the disciplinary coverage required by the research question. A problem spanning education, computing, psychology, and health, for example, may require databases or indexes with different subject coverage rather than one familiar platform.

Cross-disciplinary search guidance similarly emphasizes that databases differ in disciplinary focus and that terminology, controlled vocabularies, search fields, and syntax need to be adapted accordingly.

Create a cross-disciplinary term map

When terminology becomes complicated, create a working vocabulary map rather than keeping terms in your head.

For each major concept, record the terminology encountered in each field, the definition attached to each term, and whether you consider the relationship equivalent, broader, narrower, overlapping, or merely adjacent.

This prevents a dangerous shortcut in interdisciplinary searching: assuming that everything retrieved through neighboring terminology represents the same construct.

Do not confuse synonyms with related concepts

This distinction deserves particular attention.

“Trust,” “reliance,” “acceptance,” and “continuance intention,” for example, may all appear in literature about technology use. That does not make them synonyms. A person can trust a system without intending to continue using it, use a system despite distrusting it, or accept a recommendation for reasons unrelated to generalized trust.

Search broadly enough to discover conceptual neighbors, then read definitions closely enough to keep them distinct.

Search backward and forward from anchor papers

Keyword searching is only one route through a literature.

Once you identify a highly relevant paper, examine its reference list to find earlier work and use cited-by searching to locate later studies that build on, challenge, or apply it. This is particularly valuable when terminology has changed over time or when relevant papers do not use the words you would have searched.

Validated search research in interdisciplinary domains has shown why this matters: relevant articles may lack recognizable terminology in their titles, abstracts, or author keywords and become discoverable through indexing or citation-based approaches instead.

Search for mechanisms and outcomes when labels fail

If you still cannot find the literature you suspect exists, move below the construct label.

Search for what the phenomenon does, what causes it, how it is observed, and what outcomes it produces. Researchers may disagree about the name while describing remarkably similar processes.

For example, instead of searching only for “AI overreliance,” you might search combinations involving acceptance of incorrect recommendations, failure to verify automated advice, deference to automation, or decision errors following automated recommendations.

This behavioral translation can uncover literature whose conceptual label differs from yours.

Cross-disciplinary searching can change the research gap

A better search does more than increase your reference count. It can change the intellectual structure of the project.

You may discover that a supposedly new phenomenon has a long history elsewhere. What appeared to be a gap between disciplinary literatures may turn out to be a terminology problem. Alternatively, you may find that another discipline already contains part of the answer your field has been seeking.

Those are not failed searches. They are precisely what a literature review is supposed to discover.

Searching for Student Overreliance on Generative AI Across Fields

The results are heavily concentrated in recent educational literature. Rather than concluding that the phenomenon is new, the researcher rewrites the problem behaviorally:

People receive advice or recommendations from an automated system and sometimes follow them despite evidence that the output may be incorrect.

That description opens several possible disciplinary routes.

The researcher then compares definitions rather than merging every term into one giant synonym list. Some concepts are excluded because they concern different phenomena. Others reveal theories and evidence that substantially change the understanding of the original question.

The result is not merely a larger bibliography. It is a more defensible map of what is already known and where an interdisciplinary question might actually contribute.

Common Search Mistakes Across Disciplinary Boundaries

Build a Vocabulary Map Before You Decide the Literature Is Missing

When your research question crosses fields, treat terminology discovery as part of the intellectual work of reviewing the literature.

The aim is not to produce the longest possible list of search terms. It is to construct a defensible translation between the research problem and the different scholarly languages through which that problem has been studied.

Before Finishing a Cross-Disciplinary Literature Search, Check This

Questions About Searching Across Disciplinary Terminology

Search the Problem in More Than One Scholarly Language

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.