Research data management
Introduction
Research data include all material that supports the findings of your research question or underpins your research publications, in any form eg, digital/physical. Every researcher will work with some sort of research data, which could range from photographs of archival documents, video files, bibliographies and reading notes, to spreadsheets, code and 3D models.
Research data management refers to the practice of actively looking after your research data, including effectively and safely managing, storing and organising your data during and after a research project.
Service outline
We provide a holistic set of services to assist with the management of research data at all stages of the research process, drawing on external and internal support and policies:
- advice, guidance and training covering the broad range of research data management aspects, including funder requirements and open data
- a data management plan review service
- support for data sharing agreements, including advising on secure data access options
- infrastructure for secure data access including our ESRC SafePod, Safepoint Hub and the secure data suites
Working with data
The first step in effective research data management is to complete a data management plan (DMP). A DMP is a structured document that helps you consider your data handling and outline key information about your research project.
Data management plans can differ in structure and content, but most will include descriptions of: your data and methods; mitigations for sensitive data; anonymisation procedures, informed consent and copyright; data storage and security, including for collaborations; data archiving and sharing; and, data description and documentation.
Use our online tool, DMPOnline to help you write your DMP. We have created tailored templates for Undergraduate and Masters students, PhD researchers, research staff, and for secure data use. There are standard templates for the major UK research funding bodies. To start writing your plan in DMPOnline, you will need to create an online account, then create a plan and choose the relevant template for you.
To improve your data management plan and receive advice tailored to your project, submit your data management plan to us for review. Our service is primarily for PhD Students and research staff, however in some circumstances, you may be asked to use our service if you are an Undergraduate or Postgraduate Taught course student. Share your plan for review with the data library via email or directly through DMPOnline to receive feedback within 10 working days.
Funder policy
Many research funders expect research data management to be considered before, during and after a research project and have data management or data sharing policies that grantees are required to follow. This will often involve asking applicants to write a data management plan when applying for a grant, and where possible, sharing your data, whole or in part via a data repository.
Check your funder-specific guidance:
ESRC (Economic and Social Research Council)
European Commission (Horizon Europe)
AHRC (Arts and Humanities Research Council)
EPSRC (Engineering and Physical Sciences Research Council)
MRC (Medical Research Council)
NERC (Natural Environment Research Council)
NIHR (NHS - National Institute for Health Research)
STFC (Science and Technology Facilities Council)
All UKRI funded publications must be accompanied by a data access statement which outlines where the underlying data can be found in accordance with the UKRI Open Access Policy. Where there are reasons to limit data sharing, these should be included in the data access statement.
Institutional policy
Researchers should also follow the LSE Research Data Management Policy and can use the Research Data Toolkit for further guidance on best practice.
Key points include:
- Primary responsibility for design and implementation of effective research data management lies with the Principal Investigator.
- Data Management Plans are mandatory for funded research projects, projects where a commercial data supplier has requested a plan, projects involving individual level microdata or ‘secure data’, internal LSE data and where requested by the REC.
- All data will be stored, processed and managed within the boundaries of LSE's Information Security Policy.
- When sharing data, you should ensure it is made as open as possible, but take into account any contractual, legal, commercial and ethical obligations.
Journals policy
Many publishers now have data policies for their journals, where they may ask for data to be open for peer-review, including a data access statement or for supporting data to be available openly. Example policies include PLOS, Royal Society, Springer Nature, BMJ Open, SAGE, Taylor and Francis.
It’s crucial to actively manage your research data once you start collecting it, to keep it safe from technical failures, maintain data integrity and to ensure you keep on top of legal and regulatory requirements.
Research tools: You should consider your research tools, i.e. software and hardware carefully, ensuring that all equipment or software used meets LSE’s minimum standards. The Digital Skills Lab can help with a range of specialist research tools that are supported by the School.
Storage: We recommend storing your research data in the LSE managed O365 environment, either in OneDrive (for personal projects) or in Teams or on a SharePoint site (for collaborative projects).
When working in the field, you may be required to use portable storage devices, eg, mobile phones, audio recorders, USB sticks. Where this is the case, ensure you encrypt data files on devices and transfer data to secure, managed, storage as soon as you are able.
Working with personal data: Personal data includes information such as occupation, sex, biometric data, audio/visual information e.g. interview recordings, social media data eg, handles. It’s important to understand if you are collecting this type of data, so you can fully comply with data protection legislation and ethical standards.
Most personal data in research is anonymised before further dissemination as shared datasets. To do this properly you must consider indirect as well as direct identifiers such as contextual clues, or where two bits of information, when taken together, can identify a participant. It’s also important to ensure you gain informed consent for additional use of participants’ data, including sharing anonymised data via a data repository.
Copyright: Where research data is newly created, the staff or student creating the data is the original copyright holder. When you are using third-party copyrighted material, you will need to ensure you have a legal basis for use, and any permissions for further data sharing. In many instances, data can be used and copied for non-commercial teaching or research purposes, or criticism or review, providing that the data source, distributor and the copyright holder are acknowledged.
When sharing your research data you should also think carefully about how you will license it – for more information on this see our open data page.
Documentation: Having accompanying documentation to your data can provide important contextual, methodological and technical information to help understand your data and how it’s been created.
As well as being useful to ensure data creation is consistent when working in a research team, it’s crucially important where you are planning on sharing your data after the project, to help users coming fresh to the data navigate and be able to reuse it. The UK Data Service have great guidance on data documentation at both study and data level.
Organisation: Working with data throughout a project can be messy, with multiple processes, versions, and possibly people involved, so properly organising your data can really help save time when working with your data. The UK Data Service has good examples of best practice.
Having clear file naming schema helps uniquely identify a file or set of files. File names should reflect the file content, for instance data type, collection date, version.
Equally, storing data in well-organised folder structures will help make finding data easier. Think about the hierarchy of folders and how you want to divide the data eg, by project phase, collection method etc.
Data validation and quality assurance: It’s important to have specific procedures embedded in your methods that act as quality controls to ensure data is high-quality and valid.