Data Stewardship vs. Data Governance: Roles and Responsibilities in Research

Research teams in bioinformatics frequently use the terms data governance and data stewardship interchangeably, yet the two functions occupy distinct positions in the research data management lifecycle. Data governance establishes the decision rights, policies, and accountability structures that determine who can make decisions about data and under what conditions. Data stewardship executes those decisions through daily management activities that keep data findable, accessible, interoperable, and reusable. For a research team, the practical difference matters when assigning responsibilities: governance answers who decides, while stewardship answers who acts. This article defines both roles, compares their responsibilities in research settings, and provides a role definition matrix that teams can adapt when assigning data management duties.

Defining Data Governance in Research Contexts

Data governance in research refers to the framework of policies, standards, and decision rights that determine how data can be collected, stored, shared, and used. Governance operates at the level of institutional authority and establishes the rules that all research activities must follow. In health research, governance defines who makes decisions on behalf of whom and how data can and should be used, particularly when the data involves Indigenous peoples or other communities with collective interests in data [7]. Governance structures also address the sensitivity of patient information and implement measures to safeguard data privacy and security while recognizing and mitigating bias and promoting ethical use [5].

Governance frameworks in research settings typically include several components. Institutional ethics review establishes the boundaries for what research can be conducted with human data. Socio-political dynamics shape how communities participate in decisions about their data. Data management and data stewardship functions operate within the governance structure, and overarching influences such as legislation and funding requirements constrain what is possible [7]. A governance framework also defines the conditions needed to share, link, and use health data responsibly, with public trust and engagement ensuring that stewardship reflects public values, needs, and expectations [10].

For bioinformatics research, governance decisions include determining which datasets can be deposited in public repositories, what consent requirements apply to secondary use of genomic data, and which researchers have authorization to access controlled datasets. The National Institutes of Health Genomic Data Sharing Policy provides an example of a governance instrument that sets expectations for how genomic data from NIH-funded research should be shared [3]. Research teams must align their local governance practices with such external policies while also establishing internal decision processes.

Defining Data Stewardship in Research Contexts

Data stewardship encompasses the operational activities that implement governance decisions and maintain data quality throughout the research data lifecycle. Stewardship involves the day-to-day management of data assets, including documentation, quality control, metadata creation, and preparation of data for sharing and reuse. The FAIR Guiding Principles provide a foundation for effective research data management and stewardship, emphasizing that data should be findable, accessible, interoperable, and reusable [4]. Stewards apply these principles through concrete actions such as assigning persistent identifiers, creating rich metadata, and documenting data provenance.

The shift from data ownership to data stewardship enables responsible data management in compliance with regulations [6]. This framing recognizes that researchers who generate data act as custodians instead of owners, particularly when the data involves human participants or community resources. Stewardship responsibilities extend beyond the active research phase to include long-term preservation and responsible reuse. In community-engaged research, strong stewardship of project resources and agreements protects community interests and represents an ethical imperative for communities to adopt [9].

Data stewardship competences have become a recognized professional domain, with efforts underway to define the knowledge and skills required for these roles. The FAIR data management principles and data stewardship provide a foundation for effective research data management, and data skills should be more widely included in university curricula to prepare researchers for these responsibilities [22]. Stewardship in practice requires technical skills in data curation, familiarity with repository requirements, and the ability to document data in ways that support reproducibility.

Comparing Governance and Stewardship Responsibilities

The distinction between governance and stewardship becomes clear when examining their respective responsibilities in a research project. Governance establishes the rules, while stewardship executes them. Governance decisions are typically made by institutional bodies, ethics committees, and principal investigators. Stewardship activities are performed by data managers, bioinformaticians, and research assistants who handle data on a daily basis.

Governance responsibilities include setting data sharing policies, determining access controls, establishing data retention schedules, and defining consequences for policy violations. Stewardship responsibilities include implementing those policies through concrete actions such as de-identifying datasets, formatting data for repository submission, maintaining version control, and responding to data access requests. Governance requires authority and decision rights. Stewardship requires technical competence and attention to detail.

The relationship between the two functions is hierarchical in practice. Governance creates the conditions under which stewardship operates. A research team cannot practice effective stewardship without clear governance direction on issues such as data sharing restrictions, consent requirements, and security standards. Conversely, governance without stewardship produces policies that are never implemented. The integration of human and machine intelligence in fields such as radiology highlights the need for data governance measures to safeguard data privacy and security while also requiring stewardship practices that make large imaging databases accessible for training and validating algorithms [5].

At a Glance: Governance and Stewardship Comparison

Governance Structures in Research Institutions

Research institutions implement data governance through formal structures that assign decision rights and establish accountability. These structures vary by institution but typically include committees, policies, and procedural frameworks that guide data-related decisions.

Institutional Review and Ethics Committees

Institutional review boards and ethics committees represent a primary governance mechanism for research involving human data. These bodies review research protocols to ensure that data collection and use comply with ethical standards and regulatory requirements. In health research, institutional ethics operates as one of five components that require consideration in the governance of research data pertaining to Indigenous people, alongside Indigenous governance, socio-political dynamics, data management and stewardship, and overarching influences [7]. Ethics review determines what data can be collected, how participants must be informed, and what uses of the data are permissible.

Data Access Committees

Data access committees govern who can access controlled datasets and under what conditions. These committees review data access requests, verify that applicants have appropriate qualifications and infrastructure, and ensure that proposed uses align with participant consent and institutional policies. In genomic research, data access committees play a critical role in balancing the benefits of data sharing against privacy and sovereignty concerns. Legal constraints such as data localization provisions in privacy laws increasingly challenge traditional models of data sharing, requiring governance mechanisms that can accommodate these restrictions [12].

Community Governance Arrangements

Research involving Indigenous peoples and other communities with collective interests requires governance arrangements that extend beyond institutional review. Community-engaged research projects with tribal government or health board approval show increased likelihood of community control of resources and stronger stewardship of project resources and agreements [9]. These governance arrangements may include tribal review processes, community advisory boards, and data sharing agreements that specify how data can be used and who must approve publications. Data governance for Indigenous peoples requires defining who makes decisions on behalf of whom and how data can and should be used [7].

Stewardship Activities in Bioinformatics Research

Data stewardship in bioinformatics encompasses a range of technical and organizational activities that maintain data quality and usability throughout the research lifecycle. These activities require specialized knowledge of data formats, repository requirements, and analytical workflows.

Metadata Creation and Documentation

Stewards create and maintain metadata that makes datasets discoverable and interpretable. Metadata includes information about data collection methods, sample processing, analytical parameters, and data quality. The FAIR Guiding Principles emphasize that data should be findable and accessible, which requires persistent identifiers and rich metadata [4]. In practice, stewards must decide what metadata to capture, what standards to follow, and how to structure documentation for different audiences.

Data Quality Control

Quality control activities verify that data meet expected standards for accuracy, completeness, and consistency. Stewards develop and apply quality checks at multiple points in the data lifecycle, from raw data acquisition through processed results. In genomic research, quality control may involve assessing sequence read quality, verifying sample identity, and checking for contamination. Data governance rules are necessary to guarantee sustainable operation of research data management pipelines, particularly when multiple professions and sites are involved [23].

Data Preparation for Sharing

Preparing data for sharing requires stewards to format data according to repository requirements, create appropriate documentation, and ensure that data can be reused by others. This work includes de-identification of human data, conversion between file formats, and validation that data files open and read correctly. The FAIR Guiding Principles provide a framework for these activities, emphasizing that data should be interoperable and reusable [4]. Stewards must also understand the governance constraints that apply to specific datasets, such as consent restrictions or data sharing agreements.

Version Control and Provenance Tracking

Stewards maintain version control systems that track changes to datasets and analytical scripts over time. Provenance information records how data were generated, transformed, and analyzed, which supports reproducibility and enables researchers to trace results back to their source data. Research data as well as applied methods tend to branch out into numerous intermediate and output data objects, making it difficult to reproduce research results without careful provenance tracking [23].

Governance and Stewardship in Genomic Data Sharing

Genomic data sharing presents particular governance and stewardship challenges because genomic data are sensitive, identifiable, and valuable for secondary research. The National Institutes of Health Genomic Data Sharing Policy establishes expectations for how genomic data from NIH-funded research should be shared, including requirements for data deposition, access controls, and responsible use [3]. Research teams must navigate these requirements while also addressing the specific concerns of research participants and communities.

Consent and Secondary Use

Governance structures must ensure that data sharing complies with participant consent. Consent for genomic research may specify permitted uses, restrictions on data sharing, and conditions for future contact. Stewards must implement these consent terms through technical controls that restrict access and through documentation that records consent boundaries. The governance of genomic data must also address the interests of communities, particularly Indigenous peoples who may have collective rights and interests in data about their communities [7].

Federated and Distributed Approaches

Traditional models of centralized data sharing face increasing challenges from legal constraints and ethical imperatives around privacy and sovereignty. Data visiting, where analysis occurs within the provider's computing environment without moving the data, offers an alternative that can address these concerns [12]. Governance frameworks must accommodate these distributed approaches, which require different stewardship practices than centralized repositories. Federated data access can facilitate research or large-scale quality monitoring while preserving data privacy [8].

International and Cross-Jurisdictional Considerations

Genomic research increasingly involves data from multiple jurisdictions, each with its own legal and ethical requirements. Governance structures must address these cross-jurisdictional considerations, including data localization provisions and varying standards for data protection. A pan-Canadian health data stewardship framework and governance model illustrates how jurisdictions can work together to enable cross-sectoral and cross-jurisdictional data linkage while maintaining public trust [10]. Research teams engaged in international collaborations must understand the governance requirements that apply to each dataset they handle.

Practical Workflow for Assigning Governance and Stewardship Roles

Research teams can implement a structured process for assigning governance and stewardship responsibilities. The following workflow provides a practical approach that teams can adapt to their specific context.

Step 1: Inventory Data Assets

Begin by identifying all data assets that the research team creates, collects, or uses. This inventory should include raw data, processed data, analytical scripts, and documentation. For each data asset, record the data type, source, sensitivity level, and applicable consent or regulatory requirements. This inventory provides the foundation for governance decisions about data handling.

Step 2: Identify Governance Decision Points

Determine which decisions about data require governance authority. Common decision points include data sharing approvals, access authorization, retention periods, and security requirements. For each decision point, identify who has authority to make the decision and what information they need. Governance decisions should be documented in policies or standard operating procedures that stewards can reference.

Step 3: Define Stewardship Tasks

List the operational tasks required to manage each data asset according to governance requirements. These tasks may include metadata creation, quality control checks, de-identification, repository submission, and version control. Assign each task to a specific role or individual, and ensure that the responsible person has the necessary training and resources.

Step 4: Establish Communication Channels

Create mechanisms for communication between governance bodies and stewards. Stewards need to escalate issues that require governance decisions, such as data access requests that fall outside established policies or questions about consent interpretation. Governance bodies need feedback from stewards about policy implementation challenges and emerging data management needs.

Step 5: Document and Review

Document all governance decisions and stewardship activities in accessible formats. Review the governance and stewardship structure periodically to ensure that it remains appropriate for the research program. Changes in research scope, regulatory requirements, or community expectations may require adjustments to the governance framework.

Records and Measurements for Governance and Stewardship

Effective governance and stewardship require systematic record-keeping that supports accountability and continuous improvement. Research teams should maintain records that document both governance decisions and stewardship activities.

Governance Records

Governance records document the decisions made by governance bodies and the rationale for those decisions. These records include meeting minutes from data access committees, policy documents, data sharing agreements, and consent forms. Governance records should capture who made each decision, when the decision was made, and what information informed the decision. In community-engaged research, written agreements about data use and publishing approval represent important governance records that protect community interests [9].

Stewardship Records

Stewardship records document the operational activities performed on data assets. These records include data quality reports, metadata logs, version histories, and repository submission records. Stewardship records should enable researchers to trace how data were processed and to verify that data handling complied with governance requirements. Data quality metrics can include measures of completeness, accuracy, and consistency that are tracked over time.

Measurement Approaches

Research teams can measure the effectiveness of their governance and stewardship structures through several approaches. The FAIR Guiding Principles provide a framework for assessing whether data are findable, accessible, interoperable, and reusable [4]. Teams can evaluate their data management practices against these principles to identify gaps and prioritize improvements. Data governance functions can also be assessed by reviewing whether policies are implemented consistently and whether data-related incidents are identified and addressed promptly.

Common Failure Patterns in Governance and Stewardship

Research teams encounter recurring problems when governance and stewardship structures are poorly designed or implemented. Recognizing these failure patterns can help teams address issues before they compromise data quality or compliance.

Governance Without Implementation

A common failure occurs when governance policies exist but are not implemented in daily practice. This pattern arises when policies are developed without input from the people who handle data, when policies are not communicated effectively, or when no one is accountable for implementation. Governance as stewardship protects community interests, but only when governance decisions are actually carried out through stewardship activities [9]. Teams should verify that governance policies are reflected in operational procedures and that stewards understand their responsibilities.

Stewardship Without Authority

The opposite failure occurs when stewards are expected to manage data without clear authority or governance direction. Stewards may face ambiguous situations where they must decide whether to share data, how to handle consent restrictions, or what security measures to apply. Without governance guidance, stewards may make inconsistent decisions or avoid necessary actions. Research teams should ensure that stewards have clear escalation paths for decisions that require governance authority.

Fragmented Data Management

Research data as well as applied methods tend to branch out into numerous intermediate and output data objects, making it difficult to reproduce research results [23]. This fragmentation occurs when stewardship responsibilities are distributed without coordination, when data standards are not applied consistently, or when documentation is incomplete. Teams should establish clear data management pipelines that define how data move through processing steps and who is responsible at each stage.

Inadequate Community Engagement

Research involving communities with collective interests in data requires governance structures that include community voices. Projects that fail to engage communities in governance decisions may face resistance, mistrust, or data sharing restrictions. Community-engaged research projects with tribal government or health board approval show increased likelihood of community control of resources and stronger stewardship of project resources and agreements [9]. Teams should identify communities with interests in their data and establish appropriate governance arrangements.

Limitations and Escalation Criteria

Research teams should recognize the limitations of their governance and stewardship structures and establish criteria for escalating issues that require additional authority or expertise.

Recognizing Limitations

Governance and stewardship structures cannot address every data management challenge. Teams may lack the technical expertise to implement certain data management approaches, the resources to maintain comprehensive documentation, or the authority to resolve cross-institutional data sharing issues. Data governance frameworks must be adaptable to changing circumstances, including new regulatory requirements, emerging technologies, and evolving community expectations.

Escalation Criteria

Research teams should establish clear criteria for escalating data management issues to governance bodies or institutional authorities. Escalation is appropriate when data handling may violate consent terms, when security incidents occur, when data access requests raise novel questions, or when community concerns cannot be resolved at the team level. Teams should document escalation procedures and ensure that all team members understand when and how to escalate.

Professional Consultation

Some data management challenges require consultation with specialized expertise. Legal questions about data sharing agreements, ethical questions about consent interpretation, and technical questions about data security may require input from institutional counsel, ethics experts, or information security professionals. Research teams should identify available expertise and establish procedures for obtaining consultation when needed.

Safety and Regulatory Context

Data governance and stewardship in research operate within a complex regulatory environment that includes privacy laws, research regulations, and funding requirements. Research teams must understand the regulatory context that applies to their data and ensure that governance and stewardship practices comply with applicable requirements.

Privacy and Data Protection

Research involving human data must comply with privacy laws and data protection regulations. These laws establish requirements for consent, data security, and data sharing that governance structures must implement. The sensitivity of patient information requires data governance measures to safeguard data privacy and security [5]. Stewards must implement technical controls that protect data from unauthorized access while enabling appropriate research use.

Research Ethics Requirements

Research ethics requirements establish boundaries for what research can be conducted and how data can be used. Institutional review boards and ethics committees review research protocols to ensure compliance with ethical standards. In health research, institutional ethics operates alongside other governance components to ensure that research respects participant rights and community interests [7].

Funding Requirements

Research funders often impose data management and sharing requirements that governance structures must accommodate. The National Institutes of Health Genomic Data Sharing Policy establishes expectations for genomic data sharing that apply to NIH-funded research [3]. Research teams must ensure that their governance and stewardship practices align with funder requirements and that data sharing commitments are fulfilled.

Frequently Asked Questions

What is the main difference between data governance and data stewardship?

Data governance establishes the decision rights, policies, and accountability structures that determine who can make decisions about data and under what conditions. Data stewardship executes those decisions through daily management activities that keep data findable, accessible, interoperable, and reusable. Governance answers who decides, while stewardship answers who acts.

Who should serve on a data governance committee?

A data governance committee should include individuals with authority to make decisions about data policy and access. Typical members include principal investigators, institutional review board representatives, data access committee members, and community representatives when research involves communities with collective interests in data. The committee should include people who understand both the research context and the regulatory requirements that apply to the data.

What qualifications should a research data steward have?

Research data stewards should have technical skills in data curation, familiarity with repository requirements, and the ability to document data in ways that support reproducibility. The FAIR data management principles and data stewardship provide a foundation for effective research data management, and data skills should be included in university curricula to prepare researchers for these responsibilities [22]. Stewards should also understand the governance constraints that apply to their data, including consent restrictions and data sharing agreements.

How should research teams handle data access requests?

Data access requests should be reviewed through the governance structure, typically by a data access committee or designated governance body. The review should verify that the proposed use aligns with participant consent, institutional policies, and regulatory requirements. Stewards implement access decisions through technical controls and documentation. Data visiting, where analysis occurs within the provider's computing environment without moving the data, offers an alternative for sharing data while maintaining control [12].

What records should research teams maintain for governance and stewardship?

Research teams should maintain governance records that document decisions made by governance bodies, including meeting minutes, policies, data sharing agreements, and consent forms. Stewardship records should document operational activities such as data quality reports, metadata logs, version histories, and repository submission records. These records support accountability and enable teams to verify that data handling complied with governance requirements.

How do governance and stewardship apply to community-engaged research?

Community-engaged research requires governance arrangements that include community voices in decisions about data. Community-engaged research projects with tribal government or health board approval show increased likelihood of community control of resources and stronger stewardship of project resources and agreements [9]. Governance as stewardship protects community interests and represents an ethical imperative for communities to adopt.

What should a research team do when data management issues exceed their expertise?

Research teams should escalate data management issues that exceed their expertise to appropriate authorities. Legal questions about data sharing agreements, ethical questions about consent interpretation, and technical questions about data security may require consultation with institutional counsel, ethics experts, or information security professionals. Teams should document escalation procedures and ensure that all team members understand when and how to escalate.

How can research teams assess their governance and stewardship practices?

Research teams can assess their practices against the FAIR Guiding Principles, which emphasize that data should be findable, accessible, interoperable, and reusable [4]. Teams can evaluate whether their data management practices support these principles and identify gaps that require attention. Teams can also review whether governance policies are implemented consistently and whether data-related incidents are identified and addressed promptly.

Related Bioinformatics Guides

• Data Sharing and Privacy in Genomic Research
• The Protein Data Bank (PDB): Archival Standards, Structural Validation Metrics, and Bioinformatics Integration Protocols
• The History and Evolution of Bioinformatics
• Bioinformatics in Agriculture and Crop Improvement
• ChIP-Seq Bioinformatics Workflows

References and Further Reading

• EMBL-EBI Training. European Bioinformatics Institute.
• NCBI Data Resources. National Center for Biotechnology Information.
• Genomic Data Sharing Policy. National Institutes of Health.
• The FAIR Guiding Principles. Scientific Data.
• Data governance functions to support responsible data stewardship in pediatric radiology research studies using artificial intelligence.. Pediatric radiology, 2022.
• Unraveling the Ethical Enigma: Artificial Intelligence in Healthcare.. Cureus, 2023.
• Indigenous and Tribal Peoples Data Governance in Health Research: A Systematic Review.. International journal of environmental research and public health, 2021.
• Data governance in radiology part II: Innovative opportunities for research, education, and clinical practice in radiology.. Current problems in diagnostic radiology, 2025.
• Enhancing stewardship of community-engaged research through governance.. American journal of public health, 2015.
• Advancing Data Linkage through Health Data Stewardship and Governance.. International journal of population data science, 2026.
• Ethical and Equitable Digital Health Research: Ensuring Self-Determination in Data Governance for Racialized Communities.. Cambridge quarterly of healthcare ethics : CQ : the international journal of healthcare ethics committees, 2024.
• Data visiting governance: a conceptual framework.. Human genomics, 2025.
• Whole genome sequencing to study SARS-CoV-2 transmission between university students and the surrounding community in Pittsburgh, Pennsylvania, 2020-2021.. 2026.
• Impact of stopping contact precautions on &lt,i&gt,vanA&lt,/i&gt, plasmid transmission, Northern California, 2021-2023.. 2026.
• Assessment of Microbial Transmission by Insect Vectors in Healthcare Settings: A Systematic Review. 2025.
• A comparative analysis of infection prevention and control guidelines across Kenya, Tanzania, and Nigeria.. 2026.
• Eleven quick tips for organizing a data cleaning challenge.. 2025.
• Genomic Characterization of Extended-Spectrum Beta Lactamase (ESBL)-Producing Klebsiella pneumoniae in Transplant Recipients. 2025.
• Developing a Roadmap for National Research Data Management Governance: Based on the Analysis of United Kingdom's Case. 2015.
• Enhancing Clinical Data Infrastructure for AI Research: Comparative Evaluation of Data Management Architectures. Journal of Medical Internet Research, 2025.
• A State-of-the-Art Review in Big Data Management Engineering: Real-Life Case Studies, Challenges, and Future Research Directions. Engineer, 2024.
• Research Data Management and Data Stewardship Competences in University Curriculum. IEEE Global Engineering Education Conference, 2021.
• Applying FAIRness: Redesigning a Biomedical Informatics Research Data Management Pipeline. Methods of Information in Medicine, 2019.
• How Research Data Management can Contribute to Efficient and Reliable Science. European Conference on Information Systems, 2018.
• Data Classification and Grading as an Information Governance Mechanism: A Review of Regulatory Practices, Sectoral Implementation, and Operationalization Gaps in China. Information Switzerland, 2026.

This article is educational and does not replace validated analysis plans, institutional policy, clinical interpretation, or specialist review.