Winners and runners-up alike?—a comparison between awardees and special mention recipients of the most reputable science award in Colombia via a composite citation indicator

The study addresses whether scientific recognition within a developing country—via Colombia’s Alejandro Ángel Escobar Prize (AAEP) and its special mentions—correlates with distinct bibliometric impact patterns and whether “semi-elite” special mention recipients differ clearly from prize awardees. Motivated by the global focus on Nobel-class elites and the paucity of evidence from developing contexts, the authors investigate Colombia’s scientific community where top global awards are rare but a prominent national prize exists. They pose two research questions: (RQ1) What are the overall bibliometric features of the Colombian scientific semi-elite (sCSE; special mention recipients)? (RQ2) Are sCSE and the Colombian scientific elite (CSE; AAEP awardees) completely distinctive and segregated groups in terms of scientific impact? The study is important for research evaluation in developing countries, where output trajectories, local relevance, and publication traditions differ from the Global North, and where inclusive awards may recognize diverse scholarly outputs beyond internationally indexed journal articles.

Building on classic and contemporary work on scientific elites (e.g., Zuckerman on Nobel laureates; Garfield’s citation perspectives; studies of age-productivity patterns, prize networks, gender disparities, and field coverage), the paper highlights two gaps: (1) limited attention to awards in developing countries and (2) neglect of near-winners/nominees (e.g., special mentions). Prior studies have used citation indicators to profile and sometimes predict global prize winners, mapped networks and award genealogies, and examined productivity, collaboration, and authorship patterns across disciplines. The authors argue that local awards such as AAEP, with inclusive submission practices and recognition of varied outputs (theses, reports, books), provide a different lens on elite recognition relative to global prizes, warranting comparative bibliometric analysis between winners and those receiving special mentions.

Data and sampling: • Source lists of AAEP awardees (CSE) and special mention recipients (sCSE) from the AAEF website (2000–2020) and AAEF book (1990–1999). Period analyzed: 1990–2020 (Colombia’s indexed international publishing grew from early 1990s). • Researcher profiles identified in Scopus (articles limited to 1996–2020 due to indexing fidelity). Researchers with only one Scopus-indexed publication were excluded. • For sCSE, 35 researcher profiles (31% of 111 total sCSE) were found in Scopus. For CSE, 41 profiles existed from prior work; 35 were randomly selected to match sCSE, yielding N=70 total profiles (35 sCSE, 35 CSE). • Exclusion of hyper/multi-authored papers: Documents with ≥10 authors (n=399) were excluded to avoid skew in Big Science fields and attribution challenges. • Reclassification by field: For each researcher, journals’ ASJC subject areas (via ISSN) were used to assign a predominant subject area; ties were randomly resolved. Methods by research question: • RQ1 (sCSE profiling): Descriptive-longitudinal analysis of total output and citations by category (Physics and Natural Sciences—PhySci; Environmental Sciences and Sustainable Development—EnvSci; Social Sciences and Humanities—SoSci), and yearly citations-per-article trends for top-three most prolific researchers per category, marking the year of the special mention. • RQ2 (comparison with CSE): Replication of Ioannidis et al.’s composite citation indicator C aggregating six metrics: (1) bulk impact: total citations (NC), h-index (H); (2) coauthorship adjustment: Schreiber hm-index (HM); (3) authorship-order adjusted impact: citations as single author (NS), as single or first author (NSF), and as single, first, or last author (NSFL). Metrics are log-transformed to 0–1 and summed to produce C. Distributions compared between sCSE and CSE (box–violin plots; Wilcoxon tests reported). Data handling: Researcher identification via full names and affiliation cross-checks; profiles cross-validated; analysis limited to Scopus-indexed items within the time window. Key descriptive tables summarize counts, annual growth, citations per article, and most frequent journals by category for both groups.

- Sample composition and education/affiliations: • Of 111 sCSE, 35 (31%) had ≥2 Scopus-indexed items. Female participation among sCSE awardees was ~24%, with highest female participation in SoSci. • sCSE educational backgrounds were concentrated in Colombia (~32%) and the US/Europe (~43%); frequent affiliations included Universidad Nacional and Universidad de Los Andes; among foreign universities, Harvard and University of Wisconsin were common. - Output and citation patterns (1996–2020): • sCSE totals: 825 documents (716 articles), annual growth 5.19%, 1320 citations, citations per article 17.75. By category (sCSE): PhySci (citations/article 17.12), EnvSci (21.01), SoSci (15.12). • CSE totals: 1502 documents (1290 articles), annual growth 1.34%, 2107 citations, citations per article 30.04. By category (CSE): PhySci (58.73), EnvSci (27.68), SoSci (3.71). • Most frequent journals: sCSE—Optics Communications (PhySci), Brittonia (EnvSci), Revista de Economía Institucional (SoSci); CSE—Physical Review A (PhySci), Chemosphere (EnvSci), Revista de Estudios Sociales (SoSci). - Citations-per-article trajectories around recognition: • No overall halo/push effect after receiving either the special mention (sCSE) or the AAEP (CSE). Many recognitions occurred after multiple earlier peaks, suggesting awards follow seasoned careers. • In sCSE PhySci and SoSci, recognitions typically came after major citation-per-article peaks; in EnvSci, some special mentions preceded later peaks. • For CSE exemplars, several highest peaks occurred before recognition (e.g., Poveda, Cárdenas), though some individuals peaked after (e.g., Rey in PhySci; Castillejo in SoSci). - Composite citation indicator (C) comparison (N=70): • Quartiles of C mixed both sCSE and CSE, indicating no clear-cut separation; sCSE appeared in upper quartiles alongside top CSE (e.g., Alejandro Gaviria, John Fredy Barrera among high C). • Box–violin comparisons showed largely non-significant group differences across most indices; one indicator comparison reached statistical significance (Wilcoxon p≈0.034), but overall distributions overlapped substantially. - Additional observations: • Only ~31% of sCSE and ~47% of CSE had ≥2 Scopus-indexed documents, reflecting AAEP’s inclusiveness (accepting theses, technical reports, books, and locally oriented outputs). • Disciplinary publishing traditions were evident: STEM fields leaned toward international English-language journals; SoSci leaned toward local Spanish-language journals. • Persistent gender disparity mirrors global patterns (e.g., ~38% women among Colombian researchers; UNESCO global ~29%).

The findings address RQ1 by profiling the sCSE as a heterogeneous group with modest representation in international indexes, strong ties to prestigious local and international institutions, and disciplinary publication traditions that emphasize both international journals (STEM) and local outlets (SoSci). Their citation-per-article histories show multiple peaks, with recognitions typically following rather than precipitating heightened citation impact. For RQ2, the composite indicator C reveals substantial overlap between sCSE and CSE, undermining the notion of two distinct, segregated impact strata. While some CSE attain world-class impact, many sCSE match or exceed CSE counterparts on adjusted metrics, especially when accounting for authorship roles and coauthorship. This suggests that in a developing-country context, inclusive national awards surface researchers whose impact is not cleanly separable by standard bibliometrics. The absence of a robust post-award halo contrasts with typical Nobel dynamics, aligning more with award patterns in fields like computer science where peaks may occur later and careers lengthen. Collectively, results support viewing local scientific (semi-)elites through an inclusive framework that recognizes diverse outputs and leadership roles, and cautions against over-reliance on bulk citation measures alone to delineate elites.

- The AAEP’s inclusive scope (accepting theses, technical reports, books, and varied disciplinary practices) broadens participation across career stages, reflected in the relatively lower proportion of Scopus-indexed outputs among sCSE and CSE. - Significant gender disparity persists among sCSE, mirroring national and global trends. - Many sCSE have degrees and affiliations from prestigious local and global institutions, likely facilitating faster idea diffusion and visibility. - No overall halo effect in citations per article was observed after special mentions (sCSE) or AAEP awards (CSE); recognition often followed established impact peaks. - There is no clear-cut division between sCSE and CSE using the composite citation indicator C; both groups intermix across quartiles, underscoring the need for nuanced evaluation frameworks in developing-country contexts. Future research directions include: extending analyses to other developing-country awards; integrating multiple bibliographic databases; applying causal designs (e.g., regression discontinuity) to assess award effects; deploying science mapping to characterize structural features; and incorporating altmetrics to gauge public discourse influence.

- Geographic and award scope: single-country focus (Colombia) and single award (AAEP) limit generalizability. - Data source: reliance on Scopus alone may underrepresent outputs in local languages, books, and non-indexed venues, particularly in social sciences and humanities. - Sample coverage: only researchers with ≥2 Scopus-indexed items were analyzed (31% of sCSE; ~47% of CSE), potentially biasing toward internationally indexed authors. - Authorship exclusions: documents with ≥10 authors were excluded, limiting insights into Big Science fields. - Design limitations: before/after award effects were explored descriptively for top prolific researchers; stronger causal inference would require designs like regression discontinuity or structural variation analysis. - Field assignment: ASJC reclassification used predominant subject areas and random assignment for multi-category journals, which may introduce misclassification.