Peer-reviewed · 2026
Research on how institutions can design work more fairly, sustainably, and humanely.
Evidence and practice
These peer-reviewed articles examine the structures that shape professional life: roles, workload, job control, equity, hybrid work, and role reform. All three Versions of Record are open access; each publication includes a free PDF link and a direct link to its publisher record.
Published in The Journal of Academic Librarianship, an international peer-reviewed journal published by Elsevier.
Peer-reviewed article · Open access
The Journal of Academic Librarianship · 52 (2026) · 103222
A role-architecture framework for redesigning academic-library workloads by connecting role ambiguity, conflict, and overload with job resources, resource loss, and workload equity.
Free, immediate download. No registration required.
Peer-reviewed article · Open access
The Journal of Academic Librarianship · 52 (2026) · 103275
A scoping review of 32 direct and closely related sources showing when hybrid work becomes a resource—and when it creates new pressures, visibility gaps, and inequities.
Free, immediate download. No registration required.
Peer-reviewed article · Open access
The Journal of Academic Librarianship · 52(5) (2026) · Article 103332
A conceptual-methodological framework for studying hybrid role reform in academic libraries, paired with an illustrative comparative qualitative design.
Free, immediate download. No registration required.
Current evidence · Revision R4
Research project: Governing Library Discovery Ranking
Explore 50 TREC-COVID topics, 200 candidates per topic and 10,000 candidate appearances, with 22 original ranking configurations and two saved text-only development profiles. The read-only viewer displays existing NIST human relevance judgments, retained rankings and metrics. It does not run a new search, rerank records, infer new scores or collect judgments.
Baseline is the default. The original Full and Guardrail configurations have lower mean relevance than Baseline. Their paired 95% intervals include zero; the benchmark establishes neither equivalence nor noninferiority and does not validate these configurations for adoption.
The two text-only development profiles apply no metadata rewards. All 50 topics were used during development, so their saved results do not provide independent confirmation and neither profile is promoted.
Reviewers and testers can explore the prototype at their own pace, with no time limit or completion deadline. No account or access code is required.
Historical demonstration: the separate September 24, 2026 OpenAlex snapshot contains 107 title queries and 10,075 candidate appearances representing 10,023 distinct works. This read-only demonstration has no human relevance judgments and cannot be compared as relevance evidence with the 50-topic TREC-COVID benchmark.
The viewer is in English. Inspection filters preserve the saved ranks and metrics; no new responses are collected.
Academic affiliations are provided for identification purposes. The views presented on this website are those of the author.
