How Do You Find Research That Your Keyword Search Missed? A Guide to Citation Chaining

What if the right papers never appear in your keyword search?

You can build a careful database search, try several combinations of keywords, and still miss relevant research. The problem is not necessarily that your search was poor. Researchers may describe the same concept with different terminology, terminology can change over time, and some relevant records may not contain the words you expected in their titles or abstracts.

Once you have found one genuinely relevant paper, however, you have another way to search. Instead of asking a database to match your words, you can follow the paper's connections to other research.

This approach is commonly called citation searching, citation chaining, or sometimes citation chasing. It can help you find earlier work that your paper cited and later work that cited your paper. In some searches, those citation relationships reveal studies that keyword searching did not.

Use relevant papers as starting points for another kind of search

How citation chaining helps you search beyond keywords

What is citation chaining?

A scholarly paper sits inside a network of citations. It points backward to earlier publications through its reference list. After publication, other papers may point forward to it by citing it.

Citation chaining uses those relationships to discover potentially relevant literature. You begin with one or more useful papers, often called seed papers or seed references, and explore the literature connected to them.

These directions answer different questions. Backward searching asks, "What earlier research helped lead to this paper?" Forward searching asks, "What happened in the literature after this paper appeared?"

Why can keyword searching miss relevant papers?

Keyword searching depends on language. You have to anticipate the terminology that authors, indexers, and databases use for the concepts you want to find.

That can become difficult when a field has inconsistent terminology, when an idea has acquired a different name over time, when several disciplines discuss similar concepts using different vocabulary, or when the concept you care about is not prominent in the title or abstract.

Citation searching works differently. The connection between two papers is based on a citation relationship rather than on whether both records happen to contain your search terms. This is one reason citation searching can be particularly useful as a supplementary method for topics that are difficult to capture with text-based searches. Guidance from the TARCiS statement recommends seriously considering backward and forward citation searching for systematic search topics that are difficult to search using terminology alone.

How do you do backward citation searching?

Start with a paper that is clearly relevant to your question. Open its reference list and screen the cited publications for promising titles, authors, concepts, methods, or studies.

You do not need to read every cited paper. Treat the reference list as another set of search results. Identify potentially relevant references, retrieve them, and evaluate them using the same inclusion, relevance, and quality considerations you would apply to literature found through a database search.

Backward citation searching is especially useful for tracing the intellectual or empirical foundations of a paper. A recent article may lead you to earlier studies, measurement instruments, methodological papers, theoretical work, or previous reviews that did not appear prominently in your original search.

How do you do forward citation searching?

Forward citation searching reverses the direction. Instead of asking what the seed paper cited, you ask which later publications have cited the seed paper.

Citation indexes and scholarly search systems can provide this functionality. Google Scholar, for example, provides a "Cited by" link for many records. Citation databases such as Web of Science and Scopus also support citation-based searching.

Forward searching can be particularly useful when your seed paper is older. It lets you see how subsequent researchers used, tested, criticized, replicated, extended, or otherwise engaged with that work.

A citation is only a lead, however. A later paper may cite your seed paper for a minor methodological point, background statement, criticism, or reason unrelated to your research question. Every result still needs to be screened for relevance.

What makes a good seed paper?

Citation chaining is only as useful as the papers from which you begin. A poor starting paper can send you through a largely irrelevant part of the citation network.

Useful seed papers are usually publications you already know are closely related to your research question. Depending on your purpose, these might include a particularly relevant empirical study, a key methodological paper, a major earlier study, or a relevant review.

You do not have to choose only highly cited papers. Citation count and relevance are different things. A less-cited paper that closely matches your population, phenomenon, method, or research question may be a much better seed than a famous paper with only a loose connection to your study.

Should you chain from one paper or several?

Usually, several well-chosen seed papers are safer than relying on a single one. Every paper occupies only part of a research literature. Starting from one paper can disproportionately expose you to the authors, theories, methods, and citation traditions surrounding that particular work.

Using seed papers that represent different authors, time periods, approaches, or disciplinary perspectives can help you explore more than one part of the literature.

Can you keep following citations from papers you discover?

Yes. Suppose Paper A leads you backward to Paper B, and Paper B turns out to be highly relevant. You can examine B's references and its citing papers as well. This iterative process is sometimes described as snowballing.

That does not mean you should follow every citation indefinitely. Citation networks expand quickly. Continue selectively, using your research question and relevance criteria to decide which newly discovered papers deserve further chaining.

Are backward and forward citation results identical across databases?

No. Citation systems differ in the publications and document types they index, their coverage across disciplines and years, and how they construct citation links. As a result, the papers shown as citing a particular article can differ among Google Scholar, Scopus, Web of Science, and other sources.

For an ordinary literature review, you may not need to compare every citation database. For a systematic or otherwise highly comprehensive search, database choice, documentation, reproducibility, and coverage deserve more deliberate attention.

Is citation chaining enough for a literature search?

Usually not. Citation searching has a different strength from keyword searching, but it also has different weaknesses.

The methods therefore complement one another. Keyword searching can give you candidate papers across a topic. Citation chaining can then explore relationships around particularly relevant papers and potentially expose literature your terminology-based search did not retrieve.

When does citation searching need to be documented formally?

If you are casually exploring literature for your own understanding, detailed documentation may be unnecessary. If you are conducting a systematic review, scoping review, or another evidence synthesis intended to use a transparent and reproducible search process, the standard is different.

PRISMA-S specifically addresses citation searching. It recommends reporting whether cited references or citing references were examined and describing the methods used to locate them. It also calls for identifying the seed references on which citation searching was performed and the relevant citation database or method.

In other words, "we also checked citations" may not provide enough detail for a rigorous evidence synthesis. Record what you searched, where you searched it, which seed papers you used, and how the process was conducted.

How one relevant paper can reveal literature your keywords missed

Common mistakes when following citation trails

Build citation chaining into your literature-search workflow

Do not wait until the end of your literature review and then randomly browse reference lists. Citation chaining is more useful when it becomes a deliberate extension of your search.

The most productive workflow is often iterative. Keyword searching finds useful papers. Those papers reveal citations and new terminology. Citation chaining identifies more papers. Those papers, in turn, may suggest better search terms or additional seed references.

You are no longer relying on one perfect search query. You are learning from the literature as you search it.

Before you finish your literature search, check the citation trails

Frequently asked questions about citation chaining

A good literature search follows both words and connections

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.