Strengthening research integrity: which topic areas should organisations focus on?

The study addresses how research performing organisations (RPOs) and research funding organisations (RFOs) can best promote research integrity (RI) across disciplines. Against a backdrop of rising concerns about scientific fraud, questionable research practices, and reproducibility, the paper highlights the importance of trustworthy research and existing international frameworks and codes for RI. Despite multiple guidance documents, there is limited evidence on effective institutional interventions and a lack of discipline-specific guidance beyond medical sciences. The authors identify a gap: little is known about disciplinary differences in RI support needs. The research question is: which RI topics would researchers and stakeholders from four main areas (humanities, social sciences, natural/technical sciences, and medical/biomedical sciences) prioritise for RPOs and RFOs? The purpose is to inform concrete, context-sensitive RI policies and actions for organisations.

The paper situates its inquiry within literature documenting RI challenges (e.g., prevalence of fraud and QRPs; reproducibility concerns) and the proliferation of international RI statements and codes (Singapore Statement, Montreal Statement, Hong Kong Principles, ENRIO guidance, ALLEA Code). A scoping review found most RI guidance is generic and aimed at RPOs, with few RFO-originated practices and limited discipline-specific materials outside medicine. Evidence for effective RI interventions remains scarce; earlier systematic reviews found limited high-quality evidence, mainly for plagiarism training, and few organisational-level successes. More recent work highlights the importance of transparent research climates and micro-organisational contexts in shaping responses to RI initiatives. Economic models suggest balancing incentives for productivity with anti-fraud policies. Prior Delphi work identified priority RI topics for RPOs and RFOs, but disciplinary differences remained unexamined, motivating the present focus group study.

Design: Qualitative focus group study exploring perceived importance and prioritisation of predefined RI topics for RPOs (9 topics) and RFOs (11 topics), derived from prior Delphi work. Each session comprised: open questions on RI support needs; discussion of selected topics; and a structured sorting/ranking exercise. Sampling and setting: 30 focus groups across 8 European countries (Denmark, Spain, Netherlands, Germany, Belgium, Croatia, Italy, Greece), conducted Dec 2019–Apr 2020. Fourteen groups focused on RPO topics with researchers only; sixteen groups focused on RFO topics with mixed participants (researchers plus stakeholders such as RI/ethics committee members, funders, RPO management, industry, publishers). Total N = 147 participants (77 female, 70 male) across humanities (7 groups, n=34), social sciences (7 groups, n=32), natural sciences incl. technical (8 groups, n=42), and medical sciences incl. biomedicine (8 groups, n=39). RPO groups were further stratified by research orientation (e.g., qualitative vs quantitative social science; experimental vs theoretical natural sciences; basic vs clinical medical sciences). Purposeful and snowball sampling aimed at diversity in discipline, seniority, gender, and stakeholder roles. Sorting exercise: Participants sorted topics into categories: very important, somewhat important, none/minimal importance (resulting heat maps use five categories due to occasional in-between placements). Laminated cards listed each topic with associated subtopics to standardise interpretation (topics and subtopics as per Table 1). Ethics: Approvals from Aarhus University REC (ref. 2019-0015957) and national approval in Croatia. Informed consent obtained; confidentiality emphasised; GDPR-compliant data management. Data collection and analysis: Sessions in English were recorded, transcribed, and coded in NVivo 12, primarily deductively against predefined topic categories, with inductive coding of emergent themes. Two-stage coding generated heat maps: first, two researchers coded each group’s placement per topic into five categories; second, discrepancies were resolved and categories scored 1–5 to compute average scores visualised as heat maps for combined and per-area results.

RPO topics (combined across 14 groups):

• Very important: Responsible supervision and mentoring; Research environment.
• Important: Education and training in RI; Dealing with breaches of RI; Research ethics structures; Data practices and management; Publication and communication.
• Somewhat important (overall): Collaborative research among RPOs; Declaration of competing interests (though each was rated very important by several groups). Disciplinary patterns for RPOs:
• Humanities: Research environment was top priority; most other topics important except Collaborative research among RPOs (lower, reflecting fewer co-authorships). Marked variability across subfields; ethics and data needs differ (e.g., linguistics vs medieval history). Call for discipline-sensitive RI policies to ensure relevance and legitimacy.
• Social sciences: Eight of nine topics rated important/very important. Research ethics structures viewed of minimal importance not due to irrelevance, but because already well-covered institutionally. Collaborative research among RPOs lower overall, with quantitative groups rating it higher than qualitative groups.
• Natural/technical sciences: All topics important/very important except Declaration of competing interests (overall somewhat important; divergent views between groups). Experimental groups rated Research ethics structures very important; theoretical group rated it not important. Common issues: authorship conflicts, publication pressure, supervision challenges, and university–industry collaboration tensions (e.g., sponsor influence on wording).
• Medical/biomedical sciences: Highest priorities were Education and training in RI, Responsible supervision and mentoring, and Research environment. Dealing with breaches of RI, Research ethics structures, and Data practices and management also important. Declaration of competing interests seen as important but sometimes viewed as a formality; Publication and communication and Collaborative research among RPOs scored lower overall, with clinical groups prioritising them more than basic research groups. Noted need for harmonisation across institutions/countries; suggestion for EU-level RI structures. Top-prioritised RPO topics per area (from Table 3):
• Humanities: Research environment; Declaration of competing interests; Education and training in RI.
• Social sciences: Responsible supervision and mentoring; Dealing with breaches of RI; Data practices and management; Declaration of competing interests; Research environment; Publication and communication.
• Natural sciences: Education and training in RI; Research environment; Dealing with breaches of RI; Research ethics structures; Responsible supervision and mentoring.
• Medical sciences: Education and training in RI; Responsible supervision and mentoring; Research environment; Dealing with breaches of RI; Research ethics structures; Data practices and management. RFO topics (combined across 16 groups):
• Important (combined): Nine of eleven topics. Intellectual property and Collaboration within funded projects were overall somewhat important but still often rated important/very important in many groups. Disciplinary patterns for RFOs:
• Humanities: Eight topics important/very important; Intellectual property and Collaboration within funded projects only somewhat important; Declaration of competing interests minimal importance. Concern about added bureaucracy.
• Social sciences: Broad support; Research ethics structures and Publication & communication were very important in all groups. Dealing with breaches of RI, Selection & evaluation of proposals, and Collaboration within funded projects very important in most groups. Warnings about administrative burden.
• Natural/technical sciences: Overall validation of all topics (Publication & communication only somewhat important in combined). Differences driven by national context (e.g., Danish legal framework reduced perceived need for funder-led processes for breaches). Emphasis on appropriate division of responsibilities between RFOs and RPOs.
• Medical/biomedical sciences: Strong emphasis on role demarcation—funders should avoid parallel, potentially conflicting procedures; better to require that institutions have adequate RI structures and policies. Declaration of competing interests and Selection & evaluation of proposals seen as core funder domains. Lower ratings (somewhat important) for Funders’ expectations of RPOs, Collaboration within funded projects, and Monitoring of funded applications due to feasibility and bureaucracy concerns. Top-prioritised RFO topics per area (from Table 4):
• Humanities: Dealing with breaches of RI; Research ethics structures; Independence; Publication and communication.
• Social sciences: Publication and communication; Research ethics structures; Selection and evaluation of proposals; Collaboration within funded projects; Dealing with breaches of RI.
• Natural sciences: Dealing with breaches of RI; Declaration of competing interests; Funders’ expectations of RPOs; Updating and implementing RI policy.
• Medical sciences: Declaration of competing interests; Research ethics structures; Selection and evaluation of proposals; Updating and implementing RI policy; Independence; Publication and communication.

The findings directly address the research question by identifying which RI topics are prioritised across and within major research areas for both RPOs and RFOs. They show that while a core set of topics is broadly valued, priorities and rationales vary by disciplinary practices, research orientations (e.g., theoretical vs experimental), and national/organisational contexts. Key implications include: (1) RI policies must be sensitive to disciplinary and methodological differences to be legitimate and effective; (2) RPOs should foreground building a healthy research environment as a foundation for RI, addressing incentive structures, power imbalances, supervision quality, and diversity; (3) RFOs should translate priority topics into clear, feasible actions in their domains (e.g., fair proposal selection, conflicts of interest policies), while avoiding duplication or interference with institutional procedures; (4) coordination and harmonisation between RPOs and RFOs can reduce confusion and bureaucratic burden; (5) national legal frameworks shape the appropriate allocation of responsibilities, requiring contextual tailoring; (6) to support uptake, RI measures should minimise administrative overhead and be practical for researchers. Overall, the results reinforce a systems-level approach to RI that aligns incentives, structures, and guidance across actors, moving from principles to tailored, implementable policies.

This multi-country focus group study provides an empirically grounded map of RI topic priorities across humanities, social sciences, natural/technical sciences, and medical/biomedical sciences for both RPOs and RFOs. It confirms broad support for established RI topic areas while revealing meaningful disciplinary and contextual differences that should shape policy design. The study offers six practical recommendations: consider disciplinary differences; RPOs should build a sound research environment; RFOs should adapt priority topics into concrete, non-duplicative actions; account for organisational and national differences; avoid unnecessary bureaucracy; and develop coherent organisational integrity plans. Future work could develop and evaluate discipline- and context-specific SOPs, explore effective mechanisms for RPO–RFO harmonisation, and assess the impact of RI interventions on research climates and outcomes over time.

Not all disciplines within each main area were represented, limiting granularity at the discipline level. The design focused on main research areas rather than specific disciplines, precluding detailed discipline-specific guidance. Although sampling strove for diversity, not all focus groups met all sampling criteria. Findings reflect perceptions within eight European countries and may be influenced by national legal and institutional contexts. As a qualitative focus group study, results indicate perceived priorities rather than causal effects of interventions.