03 · What You Need to Know
How Responsibility for Long-Term Research Evidence Is Assigned
Why Does Evidence Still Matter After a Study Is Completed?
A completed study produces more than a manuscript or final report. Its conclusions depend on observations, procedures, decisions, and analyses that may need to be examined after the original research activities have ended.
Records may be needed to respond to methodological questions, verify reported findings, investigate an error, support legitimate secondary analysis, or demonstrate compliance with institutional and sponsor requirements.
Publication does not necessarily establish that every claim is correct. Peer review ordinarily evaluates a submitted account of the research rather than independently verifying every original observation.
Preserving appropriate evidence allows later scrutiny to proceed on the basis of records rather than recollection. This is particularly important when the researchers who conducted the work are no longer available.
The obligation to preserve evidence should therefore be distinguished from the active duration of the research project. A study can be operationally complete while its records remain subject to retention and accountability requirements.
What Evidence Should Be Preserved?
The required materials depend on the research design, applicable policies, and the evidentiary role of each record.
Research evidence may include original datasets, laboratory notebooks, field notes, instrument outputs, survey records, analytical scripts, coding frameworks, data dictionaries, methodological decisions, and documentation connecting the original observations to the reported findings.
Ethics approvals, consent documentation, protocol amendments, and relevant correspondence may also be subject to preservation requirements.
Not every temporary file or duplicate working copy necessarily requires long-term retention. The important question is whether the records are required by applicable rules or are necessary to understand, verify, or defend the study.
This distinction is addressed by the broader principles concerning which research records should be retained and what constitutes an adequate research record.
What Is the Institution Responsible For?
Universities and research organizations commonly establish policies governing the custody, storage, retention, security, and authorized disposal of research records.
Depending on the institution and project, institutional responsibilities may include maintaining approved repositories, providing secure storage infrastructure, defining retention schedules, managing access, and establishing procedures for personnel transitions or research-integrity investigations.
Some institutions also retain ownership or formal custody of records generated through institutionally supported research. Others establish different arrangements through contracts, collaboration agreements, or project-specific policies.
Institutional responsibility should not be assumed solely because the institution supplied the research facilities. The governing documents must establish the actual obligations.
For federally funded research in the United States, applicable award conditions and regulations may create additional requirements. For example, 2 CFR 200.334 establishes record-retention requirements for covered federal awards, subject to specified exceptions and other applicable provisions.
That regulation does not establish a universal retention period for every research record or institution worldwide.
What Is the Principal Investigator Responsible For?
The principal investigator often has operational responsibility for ensuring that project records are created, maintained, and transferred according to the governing requirements.
Depending on institutional policy and the research agreement, this may involve identifying authoritative records, establishing data-management procedures, assigning record custodians, supervising documentation practices, and confirming that required materials remain accessible after project completion.
The principal investigator may also need to respond to questions about the research and coordinate access to supporting evidence.
However, responsibility for project oversight does not necessarily mean that the principal investigator personally owns every record or must store all materials on an individual computer.
Delegating storage or recordkeeping tasks does not automatically eliminate the investigator's responsibilities under the applicable project arrangements. At the same time, institutional duties should not be transferred informally to an individual without adequate authority, resources, or access.
What Responsibilities Do Coauthors and Other Researchers Retain?
Coauthors, research assistants, data analysts, laboratory personnel, and collaborators may hold different responsibilities for the records they create or manage.
A statistician may be responsible for documenting analytical code and model specifications. A laboratory researcher may maintain original instrument records. A qualitative analyst may preserve coding decisions and analytical memos.
These responsibilities should be clearly assigned during the project and addressed when personnel leave.
Authorship also carries accountability expectations, although the scope differs among disciplines and publication standards. Under the International Committee of Medical Journal Editors recommendations, authors agree to be accountable for the work and to help ensure that questions concerning accuracy or integrity are appropriately investigated and resolved.
This does not mean that every coauthor must personally possess every original dataset. Rather, the team should have arrangements that allow legitimate questions about the research to be addressed.
The distinction between individual contributions and institutional custody is especially important when considering ownership of laboratory notebooks, field notes, and other research records.
Why Is a Designated Record Custodian Necessary?
A record custodian is the person or organizational unit assigned responsibility for maintaining specified records and managing their authorized retrieval or transfer.
Depending on the project, this role may be held by the principal investigator, laboratory manager, institutional repository administrator, records-management office, or another authorized person.
Custodianship should include sufficient information about where the records are stored, how they are organized, who may access them, and what retention or disposal requirements apply.
The custodian does not necessarily have unrestricted authority to analyze, publish, disclose, or destroy the information.
Preservation becomes vulnerable when responsibility is expressed only as a general expectation that someone on the team will keep the files. A named role and documented transfer procedure provide a more reliable arrangement.
How Long Must Research Evidence Be Preserved?
There is no single retention period applicable to all research.
Requirements may differ according to institutional policy, sponsor conditions, applicable regulations, the nature of the records, contractual obligations, and the jurisdiction in which the research is conducted.
For example, the U.S. federal award rules in 2 CFR 200.334 generally prescribe a three-year retention period for specified award records, calculated from the relevant reporting event, but include exceptions and circumstances requiring longer retention.
That provision should not be interpreted as permission to destroy all research data three years after publication. Other requirements may apply, and the relevant starting point is not necessarily the publication date.
Clinical research, regulated laboratory studies, intellectual property matters, ongoing investigations, and contractual arrangements may involve different or extended obligations.
Researchers should identify the applicable retention schedule and any circumstances that suspend routine disposal.
Does Preserving Research Evidence Mean Making It Public?
No. Preservation, access, and public sharing are different responsibilities.
Research evidence may contain personal information, confidential interviews, sensitive locations, proprietary data, or materials governed by restricted-use agreements.
Such records may need to be preserved securely without being publicly distributed.
For example, a study involving identifiable student records may require the institution to retain controlled source data while sharing only de-identified analytical materials or appropriate metadata, if permitted.
Authorized verification may occur through restricted access or institutional review procedures.
The NIH Data Management and Sharing Policy illustrates the distinction between managing scientific data and sharing them appropriately. Its requirements apply to covered NIH-funded research and recognize the need to account for privacy, confidentiality, and other limitations.
Preservation should support responsible verification without treating unrestricted disclosure as the default.
What Happens When the Principal Investigator Leaves the Institution?
A principal investigator's resignation, retirement, or transfer should trigger a review of record custody and continuing obligations.
The institution may need to identify a successor custodian, preserve authoritative records, and establish whether the departing investigator may retain or access particular materials.
Transfer arrangements may be governed by institutional policy, sponsor approval, data-use agreements, or collaboration contracts.
The investigator should not assume that taking personal copies is permissible merely because they led the project. Equally, an institution should not assume that removing the investigator's system access automatically resolves the need for continuity.
The principles governing research record handovers when team members leave remain relevant after project completion.
Who Is Responsible When Records Are Lost?
Responsibility for a loss depends on the applicable duties, the circumstances of the incident, and the evidence concerning what occurred.
A researcher may have failed to follow an established storage procedure. An institution may have provided inadequate infrastructure or unclear custody arrangements. A technical failure may have occurred despite reasonable precautions.
These circumstances should not be treated as equivalent.
The appropriate response is to investigate the loss, preserve remaining evidence, determine whether authentic copies can be recovered, and assess the consequences for the study.
The fact that records are missing does not automatically establish research misconduct. Nor does the absence of deliberate wrongdoing necessarily eliminate administrative, contractual, or professional accountability.
Where important evidence is unavailable, the procedures for responding to missing research records should guide the immediate response.
What Makes Long-Term Preservation Effective?
Effective preservation requires more than retaining files on a storage device.
Records should remain identifiable, interpretable, protected against unauthorized modification, and retrievable by authorized personnel for the required period.
For digital materials, this may require controlled backups, integrity checks, appropriate file formats, metadata, version histories, and plans for technological obsolescence.
For physical materials, appropriate environmental conditions, inventories, controlled access, and documented custody transfers may be necessary.
Preservation also depends on contextual documentation. A dataset without variable definitions or an analytical script without its required inputs may survive physically while becoming difficult to interpret.
Maintaining a traceable research audit trail can help preserve the relationship between original evidence, subsequent changes, and reported findings.
