TRUST TRACE
Trust Research Analytics and Citation Explorer. Tracing trust across disciplines.
Public trust surveys measure how people trust. TRUST TRACE maps how trust research connects — across disciplines, countries and institutions — and is developing indicators such as a cross-field connection score. These indicators describe research activity, not public trust or research quality. Name proposed to the Board, 13 Nov 2026, subject to trademark clearance.
Historical OpenAlex recovery · October 2026 snapshot · Scope screening pending
What do these numbers count? ↗01 / THE RESEARCH LANDSCAPE
Follow a connection.
One historical collection.
Three ways to explore it.
Where research fields meet
Select a field to explore its recordsExplore the records
Filters use cached titles and metadata. AI and human trust have not been systematically classified across this recovery.
02 / A FRAMEWORK FOR CONNECTION
Explore through the TRUST Framework
See the five lenses applied to a separate pilot. Read the full framework ↗
TRUSTWORTHINESS
The revised guide will test self-trust and scientific levels separately. The original pilot has not been recoded.
received a label
Assignment is provisional; it is not a reliability or quality score.
See it in the pilot papers
The three groups were chosen for the pilot, 10 papers each. They are not population shares or exclusive meanings of trust.
Self-trust, social spheres and scientific levels
03 / FROM SELF TO SOCIETY
More than one dimension.
Social scale and scientific perspective
describe different aspects of a study.
WHERE TRUST HAPPENS
Social spheres
HOW TRUST IS INVESTIGATED
Scientific levels
TRANSPARENT BY DESIGN
Know what
you are looking at.
An experimental edition, prepared 4 November 2026. Explore the evidence and its limits together.
What do the headline numbers count?
The 25,408 records are an existing, unscreened historical OpenAlex recovery with publication years 1990–2024. Coverage in 2024 is very incomplete. This is not a complete universe of trust research or a newly defined review edition.
Countries are distinct author-affiliation codes; they do not describe study populations or where trust is highest. Each work counts once per country and research field, so categories overlap. The 214 subfields are distinct OpenAlex subfield IDs from all retained topic assignments, without an extra score threshold. They have 210 distinct names; some labels occur under more than one ID. Cross-country records have two or more affiliation countries.
What do the connections mean?
A field link joins two OpenAlex fields assigned to the same work. Wider links mean more shared records; circle area reflects record count. Positions are for readability, not a scientific distance. The overview shows the eight most frequent fields. Country bars show the 12 largest affiliation counts.
Co-assignment can help find work across disciplines. It does not demonstrate interdisciplinary integration or stakeholder co-production. Citation networks and co-authorship networks require separate definitions.
What is provisional in the framework pilot?
The separate pilot contains 30 purposively selected papers, ten in each sampling group. One AI coder used locally held titles and abstracts under guide v0.1. No independent human validation has occurred. Explicit and interpreted labels are shown separately; unassessable evidence is not a negative finding.
The five TRUST facets draw on Krueger et al. (2025), A call for transdisciplinary trust research in the artificial intelligence era, DOI 10.1057/s41599-025-05481-9. The expanded guide v0.5 adds distinctions to test, including self-trust and separate scientific levels; this preview does not silently recode the original pilot.
Where does purely technical trust belong?
The agreed future structure keeps purely technical device/network trust in a separately labelled collection. Papers connecting technical systems to human or institutional trust can bridge the collections. This policy has not yet been applied to the historical recovery shown above; it can still contain technical and lexical near-matches.
What remains separate?
The 149-work demo baseline, 172-work published review, candidate collections, 25,408-work recovery and planned reproducible edition are distinct. This preview does not merge them, estimate societal trust or calculate a composite trust score.
No new search, full-text retrieval, newsletter signup or public deployment runs from this page. All interactions stay on this computer.