The research data lifecycle is a series of stages that data moves through from the initial planning of a study to the eventual reuse of that data by other researchers. The most accurate description is a cyclical model with interconnected stages: Plan, Acquire, Process, Analyze, Preserve, Share, and Reuse. These stages aren’t strictly linear. Data often loops back through earlier phases, and certain activities like documentation and quality control run continuously across all of them.
If you encountered this question on an exam or assignment, the correct answer is almost certainly the one that emphasizes this cyclical, ongoing nature rather than describing data management as a one-time task or a simple start-to-finish sequence. Here’s what each stage involves and why the lifecycle framing matters.
The Seven Core Stages
While the exact terminology varies between institutions and fields, most lifecycle models share the same core structure. The version used by the National Network of Libraries of Medicine breaks it into seven stages:
- Plan: A researcher designs the study and identifies what data they need to collect, how it will be stored, and how it will eventually be shared.
- Acquire: Data is collected through experiments, surveys, sensors, or gathered from existing sources.
- Process: Raw data is cleaned, organized, validated, and prepared for analysis. This can include reformatting, removing errors, and integrating multiple datasets.
- Analyze: Researchers run statistical models, create visualizations, and interpret the processed data.
- Preserve: Data is backed up and prepared for long-term storage so it remains usable well into the future.
- Share: Findings are published and the underlying data is deposited in a repository where others can access it.
- Reuse: Other researchers find the data and use it for new studies, or the original team applies it to different questions.
The reuse stage feeds directly back into planning, which is why this is described as a cycle rather than a pipeline. One researcher’s shared dataset becomes another researcher’s starting material.
Why “Lifecycle” and Not Just “Steps”
The word “lifecycle” is deliberate. It signals two things that distinguish this framework from a simple checklist. First, the process is continuous. Data doesn’t reach an endpoint after publication. It circulates, gets reanalyzed, and generates new research questions. Second, certain responsibilities span the entire cycle rather than belonging to a single stage.
The U.S. Geological Survey’s Science Data Lifecycle Model makes this explicit by identifying “cross-cutting elements” that apply at every stage: describing data with metadata, managing quality, and backing up files. Documentation, for instance, must be updated at every point to reflect what has been done to the data. Without that running record, data loses its value because no one else can understand or trust it.
How FAIR Principles Fit In
The FAIR principles (Findable, Accessible, Interoperable, Reusable) are a set of guidelines that overlay the lifecycle and shape decisions at multiple stages. Making data findable means assigning unique identifiers and rich descriptions during the preserve and share stages. Accessibility requires using standard formats and secure retrieval methods. Interoperability means structuring data so it can be combined with other datasets. Reusability depends on clear licensing, detailed records of how the data was produced, and adherence to community standards.
These principles aren’t a separate process. They’re design goals that influence how you handle data from the planning phase onward. A well-written data management plan addresses FAIR considerations before a single data point is collected.
The Planning Stage Carries Legal Weight
Planning isn’t optional or informal. Under NIH’s 2023 Data Management and Sharing Policy, any researcher seeking NIH funding must submit a formal data management plan describing how scientific data will be managed and shared throughout the project. The plan covers what types of data will be preserved, what metadata will accompany it, where it will be stored, and on what timeline it will become available. NSF has similar requirements.
These plans force researchers to think through the full lifecycle before a project begins. Decisions about file formats, storage infrastructure, and sharing timelines are much harder to retrofit after data has already been collected in an incompatible format or stored without proper documentation.
Preservation Is More Than Backup
The preservation stage is often underestimated. It goes beyond simply copying files to a hard drive. Data retention requirements vary by country, funder, and field, but minimums of 10 years are common, and some domains require preservation for 25 years or longer. That timeline creates real technical challenges: file formats become obsolete, storage media degrades, and the software needed to read certain data types may no longer exist.
Best practice calls for using standard, open file formats that don’t depend on proprietary software. This keeps data readable and usable decades after it was created. Preservation also involves assigning metadata that describes the data thoroughly enough for someone unfamiliar with the original project to understand and use it.
Sharing Pays Off in Citations
The sharing stage benefits the broader research community, but it also directly rewards the researchers who do it. A cross-disciplinary analysis found that papers indicating data availability receive about 25% more citations on average than papers that keep data private. In some fields the effect is even larger: microarray-based research articles that provide access to raw data accumulate 69% more citations than those that don’t.
Sharing also enables collaboration across institutions and disciplines. When datasets are publicly available with persistent identifiers (like DOIs), other researchers can find, cite, and build on that work. This is the mechanism that closes the loop of the lifecycle, turning one project’s output into another project’s input.
Alternative Models Use Different Labels
You may encounter lifecycle models that look slightly different depending on the source. The Digital Curation Centre (DCC) model, widely used in the UK, breaks the cycle into more granular phases: Conceptualise, Create or Receive, Appraise and Select, Ingest, Curate and Preserve, Store, Access/Use/Reuse, and Transform. It also adds ongoing activities like preservation planning and community engagement.
The USGS model uses nearly the same stages as the NNLM version but adds cross-cutting elements as a formal category. Despite these differences in labeling, every major model shares the same underlying logic: data requires active management at every stage, the process is cyclical, and documentation ties it all together. When answering a multiple-choice question about the research data lifecycle, look for the option that captures this cyclical, multi-stage, continuously documented process. That’s the most accurate description.