Using Citation Counting to Optimize Journal Selection in Libraries

Every library works with a budget that never quite stretches far enough. Subscriptions to academic journals are expensive, shelf space is finite, and the list of titles a department “must have” always grows faster than the money to pay for them. So how does a librarian decide which journals to buy, keep, or cancel? One of the oldest and most reliable answers is citation counting: a method that lets the scholarly community itself reveal which journals matter most. This technique sits at the heart of informetrics, and almost a century after it was first used, it still shapes how libraries build their collections.

History and development of citation counting 🔗

Citation counting did not begin in a computer lab. It began as a practical solution to a very ordinary problem: a small college library trying to spend its limited funds well. The story takes us back to the late 1920s, when two researchers turned a simple counting exercise into the foundation of an entire field.

The Gross and Gross study of 1927 🔗

The earliest systematic use of citation counting for journal selection is credited to P. L. K. Gross and E. M. Gross, who published their landmark paper, “College libraries and chemical education,” in the journal Science in 1927. The pair worked at Pomona College in California and faced a familiar dilemma. The college could not afford every chemistry periodical, so they needed an objective way to identify the titles a chemistry student or teacher genuinely needed.

Their method was straightforward but clever. They took the Journal of the American Chemical Society, the leading chemistry publication of the day, and went through its 1926 articles counting the footnote references to other journals. The titles that appeared most often were, by this logic, the most important for a chemistry collection. The resulting priority list placed German titles such as Berichte der Deutschen Chemischen Gesellschaft and Zeitschrift für Physikalische Chemie near the top, alongside major British and American journals. What looks obvious today was a genuine innovation: instead of relying on personal opinion or reputation, Gross and Gross let the citation behaviour of working chemists do the deciding.

From manual counts to the Science Citation Index 🔗

For decades after 1927, this kind of analysis remained painfully manual. Librarians and researchers physically counted references in printed volumes, which limited how widely the method could be applied. The next great leap came from Eugene Garfield, an information scientist who imagined citations as a way not just to rank journals but to organise all of science. In a 1955 paper in Science, “Citation indexes for science,” he proposed building an index that tracked which papers cited which.

This idea became the Science Citation Index (SCI), first published in 1963 by Garfield’s Institute for Scientific Information. To help decide which journals deserved a place in this index, Garfield and his colleague Irving Sher developed the journal impact factor, which measures the average number of citations a journal’s recent articles receive. As accounts of Garfield’s work explain, this measure grew into the best-known indicator of journal influence and now anchors the Journal Citation Reports and the Web of Science. In a 1972 article, also published in Science, Garfield laid out how citation analysis could serve as a tool for journal evaluation at a scale the Grosses could never have managed by hand. The thread running through all of this is unbroken: a counting method invented to stock one college library eventually reshaped how the world measures scientific communication.

How citation counting works 🔗

At its core, citation counting rests on one assumption: when an author cites a journal, that act signals the journal carries useful, relevant information. Count enough of these signals across a body of literature, and patterns emerge that tell you which journals the community actually depends on.

The basic methodology 🔗

The process follows a few clear steps. First, you select a source set of documents, such as the articles published in a leading journal, the theses submitted to a department, or the papers produced by an institution over a period. Second, you extract every reference these documents cite. Third, you tally how often each cited journal appears. Finally, you rank the journals by their citation frequency. The titles at the top of that ranking are the ones the collection should prioritise.

This is why citation analysis has long been a favourite tool for collection development in academic libraries. Researchers have repeatedly counted citations in dissertations to identify the core journals of fields ranging from plant pathology to polymer science. A botany department’s theses, for instance, reveal which botany journals its students and faculty truly use, which is far more reliable than guessing.

Citation counting and journal importance 🔗

The deeper logic behind the method connects to a well-known principle of literature distribution. In any field, a small number of journals tend to attract the majority of citations, while a long tail of journals receive only a few. This concentration means a library can cover most of a discipline’s important communication by subscribing to a relatively small core of high-citation titles. Citation counting is the tool that identifies that core precisely, rather than by reputation alone.

Importantly, a high citation count reflects more than popularity. It indicates that a journal publishes work other researchers find worth building upon. For a librarian deciding between two competing titles with similar prices, the one cited more heavily by the relevant community is usually the safer investment. This is exactly the kind of evaluative bibliometric indicator that institutions now use to support decisions once made purely on intuition.

Benefits of citation counting for journal selection 🔗

Citation counting endures because it solves real problems that libraries face every day, especially where funds are tight and accountability is high.

The first benefit is objectivity. A ranking built from thousands of citations is far harder to dispute than one person’s preference. When a librarian has to justify cancelling a subscription to an unhappy faculty member, citation data provides evidence rather than opinion.

The second is economy. Subscription budgets in many Indian university libraries are under constant pressure, and serial prices keep rising. Citation counting lets a library spend its money on the titles that deliver the most value to its users, ensuring scarce funds support the journals people genuinely cite and read.

The third is relevance to the local community. By counting citations from a library’s own users, such as the references in its students’ theses and its faculty’s publications, a librarian can tailor the collection to actual research patterns rather than a generic global ranking. A research-focused university and a small undergraduate college will naturally arrive at different core lists, and that is precisely the point.

Finally, citation counting supports weeding and storage decisions. Journals that are rarely cited can be moved to remote storage or dropped, freeing shelf space and reducing costs without harming the collection’s usefulness.

Limitations you cannot ignore 🔗

For all its strengths, citation counting is not a perfect instrument, and using it carelessly can mislead a library as easily as it can guide it.

Coverage and database bias 🔗

Citation data is only as good as the database it comes from, and major databases do not cover all journals equally. Studies of bibliometric methods note that Web of Science focuses on selective, high-impact titles, Scopus offers broader coverage including more regional output, and Google Scholar is the widest but least controlled. This matters enormously for India. Many valuable Indian and regional journals, along with publications in languages other than English, are underrepresented in the big international indexes. A ranking drawn only from these databases can quietly push a library toward foreign journals while overlooking strong local scholarship.

Disciplinary and timing differences 🔗

Citation behaviour varies sharply between fields. As bibliometric researchers point out, citations in some disciplines accumulate within a few years, while in others they build slowly over decades. A new journal, or one in a slow-citing field such as mathematics or the humanities, may look weak on a raw count even when it is excellent. Comparing journals across fields using citation numbers alone is therefore unfair, and counts taken at different times are not directly comparable either.

What citations do not measure 🔗

A citation does not always mean approval. Authors cite work to criticise it, to follow convention, or simply because it is famous. Citation counts can also be inflated by self-citation and other manipulation. Critics of citation-based metrics warn that indicators such as the impact factor are often misused as stand-ins for quality when they were never designed for that purpose. Citation counting tells you how often a journal is used, not whether its content suits a particular reader or whether it is the best title for a specific teaching need.

The sensible conclusion is not to abandon the method but to treat it as one input among several. Citation data works best alongside usage statistics, faculty consultation, the library’s budget, and the institution’s academic priorities.

What do you think? If you were managing a college library’s journal budget, how would you balance international citation rankings against the need to support strong regional and Indian-language journals that big databases tend to overlook? And in fields where citations build slowly, what other evidence would you trust to judge a journal’s worth?

References

1. https://garfield.library.upenn.edu/papers/grossandgross_science1927.pdf
2. https://mrc-jnu.blogspot.com/2013/05/
3. https://www.science.org/doi/10.1126/science.122.3159.108
4. https://www.sciencedirect.com/topics/social-sciences/eugene-garfield
5. https://www.science.org/doi/abs/10.1126/science.178.4060.471
6. http://istl.org/01-fall/refereed.html
7. https://link.springer.com/article/10.1007/s00005-009-0001-5
8. https://journals.sagepub.com/doi/10.1177/0193841X251336839
9. https://www.intechopen.com/chapters/1181108
10. https://ost.openum.ca/files/sites/132/2017/06/HausteinLariviereIncentives.pdf