Empowering Research Through Bibliometric Insights. biblio@vietscies.com Zalo 0916 086 008
Examples

Bibliometric Analysis Examples in Economics, Business, Education and Medicine

Four hypothetical projects in economics, business, education and medicine show how to turn an idea into a research question, a search string and a fitting set of bibliometric analyses.

12 min read Updated 2026-11-21 Biblio VietSCI

Bibliometric methods work in almost every discipline, but good designs look quite different from one field to the next. Below are four illustrative projects in economics, business, education and medicine. Each one sets out a research question, a draft search string, a suggested set of analyses and what the maps could reveal.

Important Every example on this page is hypothetical and for illustration only. None describes a published study, none contains real figures, and none is linked to any real author. Actual results depend entirely on the data you collect.

A common framework for any project

Whatever the field, a rigorous bibliometric project moves through the same stages, as set out by Donthu et al. (2021) and Zupic and Čater (2015):

Define the aim

Is the question about performance (who, where, how much) or about intellectual structure (themes, foundations, trends)?

Design the search

Choose the database, search fields, time span and filters.

Select the techniques

Combine performance analysis with the science mapping methods that fit the question.

Interpret and report

The author interprets the results using domain expertise and states the data limitations.

It also helps to separate the two halves of a bibliometric study from the start. Performance analysis answers descriptive questions with tables and charts: publications per year, leading sources, institutions and countries, citation counts. Science mapping answers structural questions with networks: themes, foundations, collaboration patterns. Most published reviews use both, but the balance between them should follow the research question. A study about national research capacity leans on performance analysis; a study about conceptual development leans on mapping.

Example 1 (hypothetical): Economics and green finance

Research question: How has research on green finance and green bonds developed, which themes does it focus on, and which countries lead it?

TITLE-ABS-KEY ( "green finance" OR "green bond*" OR "climate finance" )
AND SUBJAREA ( econ OR busi )
AND DOCTYPE ( ar OR re )
  • Analyses: annual publication growth, author keyword co-occurrence, country co-authorship and co-citation of references.
  • What the maps could reveal: clusters around green bond pricing, climate policy and financial risk; the overlay might point to recent interest in sustainability disclosure.
  • Field-specific caveat: much influential economics work circulates as working papers that Scopus does not index. Flag this as a limitation.

Example 2 (hypothetical): Business and SME digital transformation

Research question: What theoretical streams make up the knowledge base of research on digital transformation in small and medium-sized enterprises, and which research fronts are active?

TITLE-ABS-KEY ( "digital transformation" OR "digitali*ation" )
AND TITLE-ABS-KEY ( "SME*" OR "small and medium*" )
  • Analyses: co-citation of references to identify theoretical foundations (for instance the resource-based view or dynamic capabilities), bibliographic coupling to identify research fronts, and a thematic map in bibliometrix.
  • What the maps could reveal: a knowledge base anchored in a small number of classic strategy works; coupling clusters around e-commerce adoption, business model innovation and leaders' digital capabilities.
  • Field-specific caveat: business terminology varies widely, so the thesaurus needs careful work.

In both the economics and business examples, consider a sensitivity check on the subject area filter. Running the query with and without it shows how many relevant papers sit in neighbouring categories such as environmental science or computer science, and helps you justify the final scope to reviewers.

Example 3 (hypothetical): Education and AI in university teaching

Research question: Which themes dominate research on artificial intelligence in higher education, and which ones are emerging?

TITLE-ABS-KEY ( "artificial intelligence" OR "generative AI" OR "chatbot*" )
AND TITLE-ABS-KEY ( "higher education" OR "universit*" )
AND SUBJAREA ( soci )
  • Analyses: keyword co-occurrence with a year overlay, a period comparison (before and after widely used generative AI tools appeared) and country co-authorship.
  • What the maps could reveal: clusters on intelligent tutoring systems, academic integrity, assessment, and staff and student perceptions.
  • Field-specific caveat: the field moves fast and many records are still in press. Record the search date and be careful when comparing the most recent year.

Example 4 (hypothetical): Medicine and antimicrobial resistance in Southeast Asia

Research question: What do collaboration networks and core themes look like in antimicrobial resistance research involving Southeast Asian countries?

TITLE-ABS-KEY ( "antimicrobial resistance" OR "antibiotic resistance" )
AND AFFILCOUNTRY ( "Viet Nam" OR "Thailand" OR "Indonesia" OR "Philippines" OR "Malaysia" )
  • Analyses: country and organization co-authorship with fractional counting, index keyword co-occurrence (standardized biomedical vocabulary), optionally supplemented by PubMed records.
  • What the maps could reveal: the balance between intra-regional collaboration and partnerships outside the region, and clusters such as surveillance, community antibiotic use and One Health.
  • Field-specific caveat: medical papers often have long author lists, so fractional counting and a cap on authors per document matter a great deal; remove generic index terms such as "human" and "article".

The same logic transfers to other disciplines. In engineering, patent data or conference proceedings may matter as much as journal articles; in the humanities, books and non-English sources are poorly covered by Scopus and Web of Science, which limits what any map can show. In agriculture and environmental science, regional journals and grey literature play a larger role. Whatever the field, ask early whether the databases you can access actually cover the literature you want to describe.

The four fields at a glance

FieldSuggested sourcesKey analysesMain challenge
EconomicsScopus, Web of ScienceCo-citation, country co-authorshipWorking papers outside the databases
BusinessScopus, Web of ScienceCo-citation, coupling, thematic mapHighly varied terminology
EducationScopus, Web of Science (SSCI, ESCI)Co-occurrence, overlay, periodsRapidly changing literature
MedicineScopus, Web of Science, PubMedCo-authorship, index keywordsMulti-author papers, generic terms

Read across the table and a pattern emerges: the data source and the main challenge shape the design at least as much as the topic does. Planning for those constraints at the start, rather than discovering them after export, is what separates a smooth project from a stressful one.

Adapting these examples to your own topic

None of the four designs should be copied wholesale. Use them as templates and adjust three things to your own question.

  • Scope of the search: start with the core concept, then decide whether to restrict by subject area, region or time. Each restriction should follow from the research question, not from a wish to reduce the record count.
  • Mix of analyses: a question about who leads a field calls for performance analysis and co-authorship; a question about how a field is organized calls for co-occurrence and co-citation; a question about change over time calls for overlays and period comparisons.
  • Size of the dataset: very small datasets, under about 100 records, rarely support stable maps. Very large ones, beyond tens of thousands, may need a narrower question or more computing time.

Before collecting data, write one paragraph that links each planned analysis to a specific sub-question. If an analysis does not answer anything you care about, leave it out. Reviewers generally prefer a focused set of well-explained maps to a long gallery of figures.

Bibliometric review or systematic review?

Authors often ask which design to choose. The two answer different questions and can be combined.

AspectBibliometric reviewSystematic review
Main questionHow is the literature structured and how has it evolved?What does the evidence say about a specific question?
Typical datasetHundreds to thousands of recordsTens of carefully screened studies
Core methodsPerformance analysis, science mappingScreening, quality appraisal, synthesis
Reporting guidanceMethodological guides such as Donthu et al. (2021)PRISMA 2020 (Page et al. 2021)

A common hybrid uses bibliometric mapping to describe the landscape and then reads a smaller, screened subset in depth. If you take that route, report both stages clearly and keep the record counts consistent between them.

Interpretation belongs to the author

The "what the maps could reveal" notes above are illustrative scenarios. In a real study the clusters may look nothing like what you expected, and that is precisely where the method earns its value. Naming clusters, connecting them to theory and drawing policy or practical implications must rest on the author's expertise and on reading the representative papers, not on the figures alone.

In practice, this means keeping a short record of how each label and claim was reached: which keywords, which papers, which numbers. That record protects you in peer review and makes the eventual Discussion section far easier to write, because the reasoning is already on paper.

Start your own project

If you already have an idea in one of these fields, the Biblio VietSCI bibliometric analysis service can support the technical side: search strategy, data collection and cleaning, VOSviewer and bibliometrix analyses, visualization and delivery of all source files. We do not write Results or Discussion sections for authors and we do not guarantee acceptance. See pricing (40% off for domestic journals and unranked international journals) or submit a request on the order page.

Frequently asked questions

Are the examples on this page real studies?

No. They are hypothetical projects that illustrate study design. They contain no real data and do not describe any published research.

Which fields suit bibliometric analysis?

Almost any field with enough publications indexed in Scopus or Web of Science, typically a few hundred records or more, can use these methods.

Should medical studies add PubMed data?

It is worth considering because PubMed covers biomedical literature well. PubMed lacks citation data, however, so it is usually combined with Scopus or Web of Science.

Can I limit a study to one country such as Vietnam?

Yes, by adding an affiliation country condition to the search string. Check the data for different spellings of the country name.

Ready to map the knowledge structure of your field?

Send us your research topic. We estimate the dataset size and recommend the right package free of charge within one business day.