Weak Correlation Between Circulation and Citation Numbers Suggests that both Data Points should be Considered when Deselecting Print Monographs
Bibliographic record
Abstract
A Review of: White, B. (2017). Citations and circulation counts: Data sources for monograph deselection in research library collections. College & Research Libraries, 78(1), 53 – 65. https://doi.org/10.5860/crl.78.1.53 Abstract Objective – To facilitate evidence-based deselection of print monographs, this study examines to what extent there are correlations between circulation data (past and future usage) and between the borrowing and citation of print monographs. Design – Collections assessment project that used a variety of data sources and techniques, including Spearman’s rank correlation coefficient, statistical analysis, and the analysis of circulation data, last-use dates, and citation data. Setting – An academic library in New Zealand. Subjects – Two ranges of books were chosen for the study: 591 (Specific Topics in Zoology) and 324 (The Political Process). From these ranges, monographs published prior to 2001 were selected as the study sample. Methods – This project relied on two data sources: circulation data from the Library’s ILS and citation data from Scopus. All data was downloaded to an Excel spreadsheet in preparation for analysis. The researcher examined call numbers, authors and editors, titles and subtitles, publication dates, circulation counts, dates of last check-in, total number of citations, number of citations from publications released in 2010 and on, and number of citations from institution-affiliated documents. Renewal data was omitted, as it did not provide evidence of additional instances of use. Where multiple copies of a specific title appeared in the data set, the researcher totalled all circulations and recorded the most recent check-in date. The researcher found that some titles in the study sample were generic and it was impossible to determine if citation data from Scopus linked to the monograph in the library collection. These titles were eliminated from the study. Once data collection was complete, the researcher calculated two additional data elements: the number of months since the last check-in date and the number of citations from items published before 2010. Data in the Excel spreadsheet was analyzed using Spearman’s rank correlation coefficient to determine the relationship between past and future usage and between circulation and citation data. Main Results – Findings indicated that circulation and citation data are highly skewed. Many monographs in the study sample had never been borrowed and had few citations, while a small number of “celebrity titles” were borrowed or cited at a much higher rate than other monographs in the same classification. Further, results indicated that historic circulation numbers are imperfect predictors of future probability that a book will be borrowed. When taking a high-level view of the collection, highly circulated books tend to be borrowed more often than average. However, when examining monographs at the title level, high circulation is more of a probability instead of a robust indicator. An investigation of whether historic citation counts serve as an indicator of future citation followed previously established trends: monographs not heavily cited in the past are less likely to be cited in the future. Findings also found a weak correlation between local-institution monograph citation counts and total citation counts. Finally, the results demonstrated a weak correlation between circulation and citation data. As a group, well-cited books are borrowed more often than others, but at the individual title level, the effect is too random for either data set to predict the other in a reliable way. As such, circulation data and citation data can not be used as a proxy for each other. Conclusion – Neither circulation nor citation data can stand as full proxies of the value of a title. However, both provide information that reflects the status of a title within the scholarly community. In this environment, citation data should be considered equally with circulation figures. Both data points measure different phenomena and the weak correlation between them suggests that both are required to inform decisions about deselecting print monographs.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.060 | 0.400 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.035 | 0.055 |
| Science and technology studies | 0.002 | 0.004 |
| Scholarly communication | 0.010 | 0.018 |
| Open science | 0.003 | 0.004 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.019 | 0.005 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".