Participation, Persistence and Influence in an Expanding Research Field: Scientific Workforce Structure in Bentonite–Cementitious Research, 1981–2026

Abstract Standard bibliometric indicators of research-field growth, such as contributor counts and publication totals, are widely used in research evaluation despite limited direct testing of whether they track recurring participation and scholarly influence rather than one-time entry. We test this distinction using bentonite–cementitious research, an interdisciplinary materials-science domain in which annual output rose from 23 documents in 2012 to 179 in 2025, analysing 1,492 English-language Web of Science Core Collection articles and reviews (1981–2026) through author productivity, full and fractionalised publication counts, Lotka-law fits estimated by ordinary least squares and maximum likelihood, institutional concentration, longitudinal affiliation trajectories, and citation influence. Of 4,752 bibliographic author records, 78.28% appeared in one publication. Productivity concentration exceeded the classical Lotka exponent under ordinary least squares (β = 2.54) and discrete maximum-likelihood estimation (α = 2.72, 95% confidence interval [2.68, 2.77]); an audited sensitivity check did not reduce concentration among the leading records examined. Institutional output was comparatively dispersed (Gini = 0.39; HHI = 14/10,000), with the leading 15 of 1,674 affiliations accounting for 8.0% of distinct-document output. Recurrence was more restricted: 72.2% of affiliations published in one year and 77.4% within one five-year period. Full and fractionalised counts produced different productivity profiles, while publication frequency and citation influence identified partly overlapping contributor groups. The Participation–Persistence–Influence framework therefore treats these as partially coupled, level-dependent dimensions that should be reported separately while disclosing author-identity uncertainty.