Preserve

How to get started 

Go to kcl.figshare.com and create an account by clicking on ‘Log in’ in the top right corner of the page. You’ll automatically be logged in through Single Sign On and an account created for you.

Once logged in, you’ll see your ‘My data’ area, where you can start the deposit of your dataset by clicking on 'Create a new item'.  

This general guide from Figshare covers the steps involved to create and publish a dataset  help.figshare.com/article/how-to-upload-and-publish-your-data and this King’s specific KORDS User Guide (pdf) from Libraries & Collections gives more detail about what to include in each metadata field and how to use the available options.

When you are ready and have ticked ‘Publish’, your dataset will be reviewed by our team before it is published, so we may get in touch with you if we have any questions.

When you publish, you’ll be required to agree to the Data Deposit Agreement (pdf).

How much space is available? 

All accounts have an initial 25GB which can be increased on request using the ‘request more storage’ link in your ‘My data’ area. If you know in advance that you will have require more than 25GB, you can request an increase at any time.

Will there be a charge for using KORDS?

Costs may be charged for large deposits in future.

Already have a Figshare account?

It’s no problem to keep that separate, but speak to us if you’d like to claim your Figshare datasets in your KORDS account.

Virus checks

• Files up to 50MB in size. KORDS will perform a virus check on all files up to 50MB in size on upload and will identify any issues. If any infected files are identified, they must be deleted.  
• Files larger than 50MB. If you are depositing files larger than 50MB, you must scan the files for viruses before uploading them. If using a KCL Standard Operating Environment (SOE) device, as described on this KCL IT Security page (internal link), you can select files and right click to use the option ‘Scan with Apex One’ to perform a manual scan. If you are using any other device, including personal devices, then you will need to ensure that an up to date antivirus product installed, and perform a virus scan of files before uploading. If you require support to do this, please contact the IT Service Desk (internal link). 

What is personal data?

As in the KCL Research Governance pages, Data Protection Law defines personal data as any information about a living person from which they can be identified, directly or indirectly.

Requirements for depositing data in KORDS

When you deposit your data in KORDS, the Data Deposit Agreement will ask you to confirm a number of things related to personal data:

• That depositing researchers have completed mandatory KCL Data Protection and Information Security elearning.
• That the research has ethical approval, including approval for data archiving and sharing. If ethical approval was not needed for your research, please get in touch with us to discuss before depositing.

• That participant information/consent documents cover data archiving and sharing of anonymised data (please be prepared to provide blank copies of these documents). If explicit consent for archiving and sharing was not sought, please get in touch with us to discuss before depositing.
• That data protection registration is up to date in the Research Ethics Management Application System (REMAS) or King’s Data Protection Register (KDPR). If this has not been done because the dataset at no time contained identifiable personal data, please get in touch with us to discuss before depositing.
• That the dataset does not identify any individual nor could it lead to the identification of any individual

The full data deposit agreement can be found here: Data Deposit Agreement (pdf).

To assist you in anonymising your data and to assess whether data incudes any identifying information, we provide links to the following resources:

1. The Information Commissioner’s Office (ICO) definition of personal data and anonymisation code of practice guidance. (The ICO is currently consulting on new guidance, a draft of which can be found on the ICO website)
2. KCL guidance, including the Motivated Intruder test
3. KCL guidance Anonymised and Pseudonymised data - A guide for researchers and Anonymisation under GDPR 
4. MRC - Identifiability, anonymisation and pseudonymisation
5. UK Data Service - Anonymisation guidance and tools
6. Open Data Institute - Anonymisation and re-identification
7. Journal Article - Hrynaszkiewicz, I., Norton, M.L., Vickers, A.J. et al. Preparing raw clinical data for publication: guidance for journal editors, authors, and peer reviewers. Trials 11, 9 (2010).

If you have any questions about anonymisation and whether your data is appropriate to deposit and share, we recommend you contact the Research Governance Office in the first instance, at rgo@kcl.ac.uk.

Controlled access

Figshare has options to restrict and control access to files and complete records. Where there is a requirement to fully restrict or provide controlled access to datasets - for example, to respect participant consent or agreements with collaborators - Figshare’s ‘Permanent Embargo’ option should be used. Please get in touch with us at research.data@kcl.ac.uk to discuss how to do this.

In cases where data may be shared under certain conditions, a data sharing agreement should be used, to confirm and formalise the conditions of access. The Research Grants and Contracts team will be able to advise you about data access agreements. If you do not have your own data access agreement, a template agreement for anonymised data may be provided.

Depositing and giving access to identifiable or otherwise sensitive or confidential data

As described above, Figshare has options to restrict and control access to files and complete records. It is ISO27001 compliant and stores data in the EU, complying with the UK General Data Protection Regulation. Our Data Protection Impact Assessment and processes include the possibility of accepting datasets in KORDS which have reason and the necessary approvals and consents to contain identifiable or otherwise sensitive or confidential information. We are taking a cautious approach to doing this and will assess any such requests on a case-by-case basis. Please contact research.data@kcl.ac.uk to discuss before uploading your data.

How do I decide which data to keep? 

There are a number of factors which can influence your decision whether to retain or dispose of your data once your research project has ended. 

Which data do you need to keep? 

Many funders expect data generated during a research project to be preserved and shared (where appropriate) beyond the lifetime of the research project. This does not necessary mean you have to keep all of the data you create or collect. Requirements will vary across funders, but as a general rule you will be expected to preserve: 

• data that supports published research papers  
• data that is needed to verify or reproduce research findings 
• data that is considered to have "acknowledged long term value" (UKRI Common Principles on Data Policy)

Just what constitutes "long term value" can be difficult to quantify and will depend upon your own understanding of the data and its potential value to others within and outside your research community. 

Some journal publishers also have data sharing policies which require researchers to retain and make available data that underpins published research (see the  Deposit your data tab above and Publish your data webpage for additional information about publisher data policies).  

The university also requires research data to be kept beyond the lifetime of the research project, though retention rates vary depending on the nature of the data and the research for which it was created. See the College Records Retention Schedule.  

Which data do you wish to keep? 

Sometimes it is clear that your data will have value beyond the lifetime of your project, e.g. if it supports published research. But it is not always easy to determine in advance which data will have long term value or significance. Data may be used for purposes other than those for which they were created or have value beyond the original discipline or research community.  

Factors to consider include: 

• Is the data vital to your research project? 
• Is the data likely to be of interest to other researchers? 
• Will the data have historical significance? 
• Are the data unique? 
• Can the data be easily replicated? 

Are there data that shouldn't be kept or isn't worth keeping?

• Is their sufficient documentation and metadata to allow the data to be understood and re-used in the future? 
• Is future reuse of the data restricted or prohibited because of obsolete or proprietary software? 
• Does the cost of preserving the data outweigh any likely reuse value or significance?  

How long should I keep my data for?

Again, there are a number of factors which will determine how long you should keep your data for: 

What does your funder require? 

Most research councils and many other funding bodies have published guidelines on how long data should kept once a project has ended. These vary from funder to funder. For an overview of funder policies and links to relevant policy documents and guidance see our webpage Funder Policies on Managing and Sharing data.  

Additional guidance: 

• University of Southampton: Funder retention requirements 

What does the university require? 

Timescales for retention and disposal of research data are included in the King's Records and Data Retention Schedule (see pages 35-37 for guidance on research data). Periods of retention vary according to how the data are classified. The Corporate Records Management team can provide further assistance if you are unsure of which category your data belongs to: records-management@kcl.ac.uk 

NOTE: Where the university's retention schedule differs from funder policy requirements, the latter takes precedent. 

Are there any legal or regulatory requirements? 

General Data Protection Regulation 

The General Data Protection Regulation requires that data holding sensitive information should not be kept for longer than is necessary or for purposes other than those for which it was collected. However, Article 89 of the Regulation does provide an exemption for personal data "that is processed for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes". Data should not be disposed of prematurely where it might damage the interests of the data subjects. 

Further information: 

• JISC - Data Protection and Research Data
• Ethics and Confidentiality 
• Secure Your Data  

The Freedom of Information Act  

If you are planning to hold on to data which you are otherwise not required to retain, keep in mind that datasets held by public authorities, including universities, are subject to requests for access under the Freedom of Information Act (2000). Access can only be withheld where one of the Act's exemptions applies.  Exceptions include personal data, information intended for publication and information subject to a confidentiality agreement such as a signed consent form. Once an FOI request has been made the data must not be deleted. Destroying or deleting information once an FOI request has been made is a criminal offence.  

Further information:

JISC - Freedom of Information

Environmental Information Regulations 

The Environmental Information Regulations (2004) gives the public access rights to environmental information held by public authorities including universities. Again, there are exceptions, including for data which contains personal information.  

Further information:

ICO - Guide to the Environmental Information Regulations 

How do I securely dispose of my data?

For guidance on records and data disposal contact the Corporate Records Management team: records-management@kcl.ac.uk. For help with disposing of hardware and media contact the IT help desk.  

Additional guidance: 

• DCC - Five Steps to Decide What Data to Keep 
• DCC How to Appraise and Select Research Data for Curation  

Choosing file formats

Choosing open file formats is important for data archiving and long-term preservation, but it is also a good idea to think about which formats you will use to store your data even before you start collecting your data as your choice of file formats will have a significant impact on whether you will be able to access your data files at a later date and also on the ability of other, future users to access the data in the long term. 

Ideally you should use formats that allow for the long term preservation and accessibility of your data, but your choice will most likely be determined by a range of factors. These might include...  

Factors to consider 

• Which formats are best suited for data collection and analysis? 
• Which formats have you and your colleagues used in the past? 
• Are there any disciplinary-specific standards or norms?  

but also... 

• Is the format open or proprietary? 
• Does the format allow for the long-term preservation and reuse of the data? 
• Is there a risk of file format obsolescence?  
• Is the format suitable for conversion? 
• Does your chosen repository or data centre place any restrictions on the types of file formats they will accept?   

Which file formats are suitable for long term preservation and future reuse? 

Open versus closed, proprietary formats 

Where possible you should store your data in open or standard rather than closed or proprietary file formats.  

With most proprietary formats the specification is privately owned and subject to restrictions due to copyright or other intellectual property rights. 

With open formats however, the code supporting the format is publicly available and free to use by anyone. This allows others to develop software that can access the files and reduces the risk of software or hardware obsolescence.  

Open formats are also more likely to be backwards compatible with previous versions and well supported by a user community over a longer period of time. 

Examples of open formats include CSV, XML, JPEG 2000, Open Office Document, Tar, ZIP.   

Some proprietary formats have become standard and are also supported by open documentation (e.g. PDF, TIFF) or are widely used and likely to be around of a long time (e.g. SPSS, MS Office software applications) and are therefore considered acceptable for short term preservation.  

Lossless versus lossy compression 

It is generally best to avoid “lossy” compressed file formats. Lossy compression has the advantage of allowing smaller file sizes but some important information might be lost. For image files, TIFF is lossless, JPEG lossy; for audio files, WAV is lossless, MP3 lossy. Lossless compression produces larger files but every bit of the original data is restored when the file is uncompressed (e.g. PNG, GIF, and ZIP files). For important files, best practice is to keep a master copy in a lossless format.  

Recommended file formats: 

• UK Data Archive - Recommended File Formats 
• Oregon State University - Data Types and File Formats 

Data conversion and format migration 

Sometimes using proprietary formats is unavoidable, particularly if they are widely used within a discipline or research community (e.g. crystallographic information files (CIF)). It may well be the case that the formats you need to use to collect and analyse your data are not always the most suitable for preserving your data beyond the lifetime of your research project. Where possible you should consider converting your data to open formats to allow long term preservation and access. However, when converting data files from one format to another it is always advisable to check for any errors or loss of information such as missing such as fonts, footer and headers, footnotes, hyperlinks, image resolution, sound quality and colour fidelity. Also, some open formats lack the functionality and formatting of proprietary formats, so there may be occasions when it is a good idea to retain copies of important data in their original format while also using open formats to enable future access and sharing.  

The UK Data Archive provide useful guidance and information on managing  quality control  for data collection, migration and transcription. 

Where should I keep my research data once the project has finished?

• Once you have decided which data to keep you will need to find somewhere to deposit your data for long-term preservation. Ideally you should deposit your data in a reputable data repository.
• Depositing your data in a data centre or repository ensures long-term preservation, continued access and maximises reuse value for future research projects.
• Many funding agencies and an increasing number of journals now expect data that supports published research to be deposited in a data repository

Which repository should I choose?

• Wellcome Trust approved repositories is a curated list of trusted repositories, organised by data type and research area (scroll down the page to point 2.2.1.)
• re3data.org is a registry of research data repositories which allows you to search by discipline
• FAIRsharing is a useful aggregator of resources on FAIR data and metadata standards, inter-related to databases and data policies

Some examples are listed here.

Domain/disciplinary specific

• e.g. Electron Microscopy Data Bank (EMDB), Archaeology Data Service, GenBank

Generic/multi-purpose

• e.g. Dryad, Figshare, Zenodo

Institutional

• e.g. the King's Open Research Data System (KORDS)

Software/code

• e.g. GitHub,Bitbucket**
• Guidance on choosing a suitable repository for software is available from the Software Sustainability Institute web site here.

*Some funders support domain or disciplinary specific repositories or data centres and expect researchers to deposit their data in these where appropriate 

For example: 

• The NERC (Natural Environment Research Council) manages seven data centres and also supports the Archaeology Data Service for science based archaeology data
• The ESRC (Economic and Social Research Council) funds the UK Data Archive which specialises in social and economic digital research data.  

**You can ensure the long-term preservation of code/software by linking your GitHub account with Zenodo. An integration between the services allows you to log in to Zenodo using your GitHub account. Guidance is available here.  

Can I include data preservation and archiving costs in my grant application? 

The UKRI (UK Research and Innovation), Wellcome Trust, and Cancer Research UK all allow for preservation and archiving costs to be included in grant applications as long as the costs are incurred or allocated before the award has ended. The UKRI's 'Guidance on best practice in the management of research data' contains information on how to include costs in a grant application (see pp.10-12).  

Creating a data management plan will help you budget for these costs in advance.