Research Integrity Risk Index (RI²) is a diagnostic, bibliometric framework designed to assess institutional exposure to selected research-integrity-related risks. Developed in 2025 by Lokman Meho (لقمان محّو), a professor and University Librarian at the American University of Beirut, RI² is presented as complementary to conventional global university rankings, which primarily emphasize publication volume and citation counts. RI² evaluates universities using three institution-level bibliometric indicators associated with integrity risk signals in scholarly publishing: retracted journal articles, publications in journals that were subsequently delisted from major bibliographic databases, and institutional self-citation patterns. The framework emerged amid broader debates over the limitations of citation- and volume-based ranking methodologies in capturing research integrity-related risks. How does RI² work?
Methodology RI² assesses academic institutions using three bibliometric indicators derived from verifiable publications and citation data.
Delisted Journal Risk (D Rate) measures the percentage of an institution's research output published in journals that were later delisted from Scopus or Web of Science following those databases' re-evaluation processes for editorial or publishing standards.
Retraction Risk (R-Rate) measures the number of retracted journal articles per 1,000 published articles over a defined publication window. Retraction data are compiled and cross-validated using multiple independent sources, including the Retraction Watch Database, MEDLINE, Scopus, and Web of Science, with filtering procedures applied to exclude duplicate records and retractions unrelated to authorship issues.
Self citation Rate (S-Rate) measures the proportion of citations to an institution's publications that originate from the same institution. Citation data are sourced from InCites, and the indicator is reported descriptively and interpreted in relation to peer institutions within the same field classifications.
All three indicators are field normalized and benchmarked against a fixed global reference group comprising the 1,000 publishing universities worldwide. Each indicator is scaled to a 0–1 range using Min-Max normalization with caps applied to extreme values, and the normalized indicators are combined using equal weighting to generate the composite RI2 score.
Based on their composite scores, institutions are assigned to one of five percentile-based risk tiers: Low Risk, Normal Variation, Watch List, High Risk, and Red Flag. Tier placement reflects relative exposure to integrity-related publishing risks and is not intended as an assessment of research quality, misconduct, or institutional performance.
Detailed methodology is available online on the official RI² website. TierPercentile RangeInterpretationRed Flag≥ 95thExtreme anomalies; systemic integrity risk High Risk≥ 90th and < 95thSignificant deviation from global normsWatch List≥ 75th and < 90thModerately elevated risk; emerging concernsNormal Variation≥ 50th and < 75thWithin expected global varianceLow Risk< 50thStrong adherence to publishing integrity norms
Output and visualization RI² website provides tabular results for the world's most publishing universities along with world maps that illustrate country-level D-Rate and R-rate, allowing examination of geographic variations in risk profiles.
Red-flagged universities by Research Integrity Risk Index (RI²)
This table lists the universities flagged as highest-risk (Red Tier) in the Research Integrity Risk Index (RI²).