Single-author study links several formal features of ecology papers with citation impact

A quantitative study of one vegetation ecologist’s publications found that journal impact, title length, author count, article length, publication year and special-feature status were associated with differences in article-level citation impact. The analysis covered 177 works, including 174 journal articles, but its single-author sample limits how broadly the findings can be generalized.

The study used publications co-authored by Jürgen Dengler and indexed in Scopus on 1 May 2024. From 189 initial entries, the author removed four duplicates, six papers published in 2024 and two unsystematic pre-2003 entries, leaving 177 observations from 54 journals and two book series. The most frequently represented journals were Journal of Vegetation Science with 25 articles, Tuexenia with 21, Vegetation Classification and Survey with 13, Phytocoenologia with 12 and Applied Vegetation Science with 11.

Citation impact was measured using Scopus’s Field-Weighted Citation Impact, or FWCI, as of 1 May 2024. FWCI compares an article’s citations with those of papers in the same year, subject area and article type; a value of 1 represents average performance by that comparison, while 2 represents twice the average. Journal-level Source Normalized Impact per Paper, or SNIP, served as a predictor. Journal articles were modeled separately from book chapters because Scopus does not assign FWCI to the latter.

After highly skewed variables were log-transformed, a global linear regression was refined until seven significant terms remained. Together, these explained 50.2% of the adjusted variance in log-transformed FWCI among the journal articles. SNIP was the strongest predictor in the multiple model and accounted for 26.8% of variance in a simple regression. The model estimated that doubling SNIP was associated with a 43% increase in FWCI.

Title length was also strongly associated with citation impact. Each additional title word was associated with an 8.4% lower FWCI, while one fewer word was associated with a 9.1% higher value. Author count was positively associated with impact: each doubling of the number of authors corresponded to a 13.9% increase in FWCI. Doubling the number of pages was associated with a 16.1% increase, and the model estimated a 6.7% decrease in FWCI for each later publication year after accounting for the other variables.

Two binary characteristics were positively associated with FWCI. Articles with a colon in the title had an estimated 48% higher impact, and articles in special features or collections had an estimated 44% higher impact. Open-access status, English versus German publication, and titles containing a dash, question mark, word play or factual statement did not have significant independent associations in the multiple model. These findings concern statistical associations within this dataset; the analysis did not establish that changing any formal feature would cause a paper’s citations to change.

The author offered possible explanations for some patterns. He noted that shorter titles may contain a more focused sequence of searchable terms, while papers with more authors may involve larger datasets, additional expertise or broader networks of readers. Special features may increase visibility through associated editorials and focused promotion. The comparatively strong positive association for longer papers was less clear, and the source presents the possibility that longer papers cover more subtopics as an interpretation rather than a tested mechanism.

Variation within individual journals was substantial. Among journals represented by at least five papers, the most-cited article generally performed at least 10 times as well as the least-cited one, except in Journal of Biogeography. The range reached 185-fold in Vegetation Classification and Survey and 76-fold in Journal of Vegetation Science. A notable example was a 2018 paper that had been rejected by Nature and was ultimately published in Nature Ecology & Evolution; at the stated measurement date, it had FWCI of 20.30, compared with a 2022 SNIP of 11.59 for Nature.

The study’s principal limitation is its scale and composition. It includes fewer than 200 works from a single vegetation ecologist, although the articles span more than 50 journals and the author reported that author number and title length were also influential in an unpublished analysis of papers in two vegetation journals. FWCI and SNIP also depend on Scopus subject classifications, and journal-level SNIP was taken from 2022 rather than each article’s publication year. The results therefore identify patterns in this author’s publication portfolio and require broader multi-author, multi-disciplinary testing before being treated as general determinants of citation impact.

Subject of Research: Statistical analysis of formal features associated with citation impact in one vegetation ecologist’s publication portfolio

Article Title: Determinants of citation impact

Article References: Dengler, J. (2024). Determinants of citation impact. Vegetation Classification and Survey, 5, 169-177. https://doi.org/10.3897/vcs.126956

Image Credits: AI Generated

DOI: 10.3897/VCS.126956

Keywords: citation impact, FWCI, SNIP, bibliometrics, title length, author count, journal prestige, open access, special features, vegetation ecology, Determinants, citation