Methodology
Every signal on OpenRLens is computed from open scholarly data using reproducible logic. No black boxes — this page explains what each signal measures and why.
Data Sources
All data is sourced from publicly available, open scholarly databases. Nothing proprietary.
Reading the signals
A high score is a positive indicator
A high score is a risk indicator
Context only — not scored in verdict
Signals are scored 0–1 and classified as High / Moderate / Low using fixed thresholds derived from the distribution across all institutions. Computed over the date window you select.
Structural Signals
Institutional Leadership Signal
Higher is betterHow often the institution holds a leading authorship role (first, corresponding, or PI-equivalent last author) on its own papers.
Research Concentration
InformationalShare of total citations held by the top 10% of recurring authors. High = a few stars dominate output.
Collaboration Dependency
Lower is betterHow much output would drop if the top-3 external partner institutions were removed. High = research is not self-sustaining.
Affiliation Stability
Higher is betterRatio of researchers active across ≥3 distinct years vs. those across ≥2 years. A proxy for faculty retention.
Output Stability
Higher is betterConsistency of annual publication volume. Computed as 1 − (std ÷ mean). Volatile output scores low.
Thematic Continuity
Higher is betterYear-on-year Jaccard similarity of the top-5 research concept labels. High = focused, stable research identity.
Funding Signals
Funding Presence
Higher is betterShare of papers that acknowledge at least one external funder. A lower bound — open databases underreport funding.
Funding Leadership
Higher is betterSame as ILS but restricted to funded papers only. Does the institution lead the research it is funded for?
Funding Dependency Differential
Higher is betterFLS rate minus ILS rate. Positive = institution leads more of its funded work than its overall output.
Lead Role on Funded Papers
Lower is betterFunded papers with no institutional lead author ÷ total funded papers. Lower is better.
Funding-Theme Alignment
Higher is betterJaccard similarity between concept labels in funded vs. unfunded papers. High = funded work stays on-topic.
Mentorship & Diversity
Mentorship Diversity Index
InformationalShare of first authors in the recent half of the window who did not appear in the earlier half. High = new researchers are being elevated.
Gender Balance Index
InformationalEstimated gender balance among institutional first authors. Name-inference only (~80–85% accuracy for Western names). Indicative.
The overall verdict
The verdict shown on each results page is a weighted composite of the structural and funding signals. MDI and GBI are informational and do not contribute to the score. Positive-polarity signals contribute positively; negative-polarity signals contribute inversely. No single signal dominates — a Strong verdict requires broad coverage across both structural and funding dimensions.
Exact weighting coefficients are not published — not to obscure the method, but because publishing fixed numbers creates incentives to optimise the score rather than genuine research quality.
Known limitations
Question about a specific signal?
info@openrlens.com →