What Really Counts as Research Data?
Defining Research Data
Research data are the evidence that underpin research findings. However, defining exactly what counts as research data is not always straightforward. Different disciplines, organisations and policy frameworks use slightly different definitions, reflecting the wide variety of data produced across research fields. In addition, relatively little attention has been given to defining the concept itself, meaning that there is no single universally accepted definition of research data (Reichmann et al., 2021; Gomez-Diaz & Recio, 2022).
This guide adopts the definition developed by the Organisation for Economic Co-operation and Development (OECD), one of the more widely adopted definitions in research data management (Gomez-Diaz & Recio, 2022). The OECD defines research data as:
"Factual records (numerical scores, textual records, images and sounds) used as primary sources for scientific research, and that are commonly accepted in the scientific community as necessary to validate research findings." (Pilat & Fukasaku, 2007)
Put simply, research data are any materials that are collected, observed, created or generated during a research project and used as evidence to answer a research question or support research findings.
Although research data may be digital or physical, this guide focuses primarily on the management of digital research data, as these are most commonly stored, shared, preserved and reused using digital repositories and online research infrastructures.
Research data exist in every discipline
The examples below illustrate the diversity of research data across fields but are not intended to be exhaustive.
| Discipline | Examples of research data |
|---|---|
| Life Sciences | DNA sequences, laboratory measurements, microscopy images, field observations, laboratory notebooks |
| Physical Sciences | Experimental measurements, telescope observations, sensor readings, simulation outputs |
| Social Sciences | Survey responses, questionnaires, interview transcripts, census data, social media datasets |
| Humanities | Historical manuscripts, archival documents, digital editions, field notes, annotated texts |
| Law | Legislation, court judgments, legal databases, case annotations |
| Psychology | Behavioural observations, cognitive test results, questionnaires, interview transcripts, interview recordings |
| Engineering | Experimental measurements, sensor data, simulation outputs, 3D scans |
| Environmental Sciences | Satellite imagery, biodiversity records, GIS datasets, weather observations |
| Computer Science | Benchmark datasets, simulation outputs, machine learning datasets, computational results, log files |
| Arts and Creative Practice | Photographs, performance documentation, audio recordings, video recordings, digital artworks, exhibition documentation, digital images |
These examples illustrate that research data can take many different forms, including text, images, audio and video recordings, software, databases, models, measurements and physical specimens. What qualifies them as research data is not their format, but their use as evidence to answer a research question.
What counts as research data?
Although research data exist in every discipline and can take many different forms, not everything produced during a research project is research data. In reality, whether something is considered research data depends less on what it is than on how it is used.
Research data are materials that are collected, observed, created or generated to answer a research question or support research findings. By contrast, documents created solely to manage, administer or communicate a project are generally not research data.
The same type of material may therefore be research data in one project but not in another.
Materials usually considered research data
| Material | Explanation |
|---|---|
| Experimental measurements | Primary evidence supporting research findings. |
| Survey responses | Used as evidence to answer the research question. |
| Interview recordings and transcripts | Analysed to identify themes, behaviours or experiences. |
| Field observations and field notes | Record observations that are analysed as part of the research. |
| Laboratory measurements | Primary evidence supporting research findings. |
| Images, photographs or videos collected for analysis | Used as evidence to investigate or document the research subject. |
| Historical manuscripts, archival documents and annotated historical texts | Serve as primary sources in historical and humanities research. |
| Satellite imagery | Analysed to investigate environmental or geographical phenomena. |
| Simulation outputs | Generated and analysed to answer research questions. |
| Source code used to analyse data or produce research results | May be considered research data or, more commonly, research software that forms part of the research record. |
Materials usually not considered research data
| Material | Explanation |
|---|---|
| Grant applications | Prepared to obtain funding rather than answer the research question. |
| Project budgets | Used to manage the project rather than conduct the research. |
| Meeting agendas and minutes | Administrative records rather than research evidence. |
| Contracts and procurement documents | Support project administration rather than the research itself. |
| Internal emails about project management | Used to coordinate the project rather than generate research evidence. |
| Travel claims and expense reports | Financial records, not research evidence. |
| Promotional material or project website content | Created for communication rather than research. |
| General teaching materials created for a course | Developed for teaching rather than as evidence for research. Unless they themselves are being analysed as part of the research. |
Note: There are exceptions. A grant proposal, project budget, meeting minutes or teaching material could become research data if it is itself the object of a study—for example, research on science policy, research funding, higher education or organisational decision-making.
What about other research outputs?
Some materials do not fit neatly into either category and may be regarded as different types of research outputs depending on their role in the project.
| Material | Usually considered | Explanation |
|---|---|---|
| Source code used for data analysis or to produce research results | Research software | May accompany a dataset as part of the research record and, in some disciplines, may itself be analysed as research data. |
| Scripts | Research software | Used to automate analyses or workflows. Like source code, they may accompany research data but are generally treated as research software. |
| Algorithms | Research method or research software | Algorithms describe methods or procedures rather than evidence. They are usually documented as part of research software or methods, although their implementation may be shared alongside research data. |
| Computational models | Research software or research model | The model itself is generally not research data, whereas the inputs, parameters and outputs may be. |
| Laboratory notebooks | Research record | Often contain observations, measurements and experimental records that may themselves constitute research data. Electronic laboratory notebooks are increasingly preserved alongside datasets. |
| Protocols, workflows and standard operating procedures | Research methods | Describe how research is conducted rather than providing evidence to answer the research question. |
| CAD models and engineering designs | Research output | May be research data when analysed as evidence, but are often treated as design outputs or engineering artefacts. |
| Physical specimens (e.g. biological samples, archaeological artefacts, geological cores) | Research material | The specimens themselves are research materials, while their associated measurements, images and metadata are usually the research data managed and shared. |
| 3D models and digital reconstructions | Research data or research output | In fields such as archaeology, architecture and cultural heritage, they may be analysed as evidence or shared as research outputs depending on their role. |
| Spreadsheets and databases | Container or file format | A spreadsheet or database is not inherently research data. Whether it contains research data depends on the information it stores and how it is used. |
Key Takeaways
- Research data are materials used as evidence to answer a research question or support research findings.
- Research data exist in every discipline. They may take many forms, including measurements, images, texts, audio and video recordings, software outputs and physical specimens.
- Research data are defined by their purpose, not their format. The same type of material may be research data in one project but not in another, depending on how it is used.
- Not everything produced during a research project is research data. Administrative documents, financial records and project management materials generally support the research process but are not themselves research data.
- Some research outputs are distinct from research data. Research software, laboratory notebooks, protocols, computational models and physical specimens may play an essential role in research but are often managed as different types of research outputs.
References
Foundational Documents
Pilat, D., & Fukasaku, Y. (2007). OECD principles and guidelines for access to research data from public funding. Data Science Journal, 6, OD4-OD11. https://doi.org/10.2481/dsj.6.OD4. Licence: CC BY 4.0
Research Evidence
Gomez-Diaz, T., & Recio, T. (2022). Research Software vs. Research Data I: Towards a Research Data definition in the Open Science context. F1000Research, 11. https://doi.org/10.12688/f1000research.78195.2. Licence: CC BY 4.0
Reichmann, S., Klebel, T., Hasani-Mavriqi, I., & Ross-Hellauer, T. (2021). Between administration and research: Understanding data management practices in an institutional context. Journal of the Association for Information Science and Technology, 72, 1415 - 1431. https://doi.org/10.1002/asi.24492. Licence: CC BY-NC-ND 4.0
Research Infrastructures and Resources
Organisation for Economic Co-operation and Development (OECD). https://www.oecd.org/
Author: Jonathan England
Created: August 2026
