There are lots of decisions to make before you start to create your data. Making these choices early on in your project can save you time and effort later, and your decisions will affect how you can use, share and publish your data.
Many funders now expect you to show you've engaged in data planning. You can do this by writing a data management plan, which is a document describing how your data will be handled both during a research project and after the project has ended.
If you are applying for research funding you may be required to submit a data management plan as part of your grant application..
Other common names for a Data Management Plan include:
Throughout these web pages, we use the term "Data Management Plan" to cover all of these and similar documents.
Writing a data management plan for the first time can be challenging. However, it is worth taking the time to consider issues thoroughly and to find answers to any questions you identify as this will save you time and effort later.
The data management issues relevant to your area of research are unlikely to change rapidly, so you will be able to re-use and adapt aspects of your data management plan for future research projects.
A number of funding bodies require that data security questionnaires are completed for the projects that they fund. The Project Leader - IT Security (intranet link log in to access) can advise on technical aspects of data security for these questionnaires.
Writing a data management plan typically involves answering a series of questions about how you plan to create, describe, secure, retain and share your data.
Your plan should be concise and appropriate to the nature of your research, with more detailed plans for larger projects. You should justify the decisions you make and be prepared to implement your plan. You can also update your plan once your project has started to reflect changes in your research.
Because of the diversity of research, there is no single correct answer to what a data management plan should cover. However, a good data management plan should typically address the following topics:
- If you're re-using existing data, what licences or terms of use will you have to comply with?
- How will new data build on and relate to existing data? Why were existing data unsuitable for re-use in your new project?
- What types of new data will you create and in what format? Did you chose these formats because they are standards in your discipline, are linked to the software or equipment you will use, or are open file formats?
- Can you estimate the size of the data you'll create? Will it be less than 500GB, around 1TB, or substantially more than 1TB? How many boxes might non-digital data fill?
- What methods will you use to capture your data and how will these ensure that your data are high quality? Will you use standard protocols, include replicates or controls, or automate data capture.
- What contextual information is needed for you or someone else to understand your data? Do you need to record methodologies, equipment settings or abbreviations used?
- How will you capture contextual information? Will this be in a 'readme' text file to accompany the data, or will you embed metadata directly in file properties or headers?
- Are there any standards that you will use? The Digital Curation Centre maintains a list of metadata standards for different disciplines.
- Where will you store your data and how will you ensure that they are backed up? Will you use University-managed data storage (intranet link log in to view) or need to set up your own back-up procedures?
- How will you secure your data? What methods will you use to restrict access to your sensitive data? Will you encrypt hardware when working off campus?
- How will you protect your research participants? Will you obtain informed consent for data retention and sharing? How will you anonymise data to safeguard the privacy of your participants?
- Which subsets of your data will you keep at the end of your project? Will you retain anonymised versions but destroy personal data and identification keys? Will you retain all of the raw data or is a processed version more suitable to preserve? Do you need to keep all intermediary files or would you only need to refer back to input files or a final version?
- How will you prepare your data for long-term preservation? Are you able to convert your data to open file formats (UK Data Archive)? What contextual information do you need to retain so that your data remain understandable and usable?
- Where will you archive your data to ensure that they are preserved and sustained for several years after your project ends? Will you submit your data to a specialist data repository/centre and if so, have you consulted them about your requirements?
- How big will your final dataset be and will there be any costs associated with archiving them, such as data deposit charges?
- Can you demonstrate that you'll plan ahead to maximise data sharing? For example, will you only share a subset of the data where informed consent was granted for data sharing?
- Are there any reasons why you would not be able to share some of your data? Would they be covered by the Data Protection Act, licence restrictions or contractual confidentiality clauses? Are there ethical reasons why data should not be released?
- When will you share your data? Will data be made available upon first publication of findings or within a limited period after the end of the project? Do you need to delay publication to allow for commercialisation or patent applications? Will you embargo your data to allow for a limited period of exclusive use?
- How will you share your data? Will you publish supplementary information alongside articles? Will the data repository where you archive your data make it available? Will you forward copies of the data upon request?
- How will you disseminate your research? Will you include a data access statement in published articles? Does your chosen method of data preservation provide a persistent URL such as a Digital Object Identifier? What licences will you assign to your data?
Many funding bodies now require data management plans to be submitted as part of grant applications, although the format and content of these plans can differ between funders.
The University of Bath has an excellent summary of funder policies which includes information on the requirements for data management plans for different funders.
The University of Bath also has excellence guidance on funder expectations for data management plans as does the London School of Hygiene and Tropical Medicine (funder requirements for data management and sharing).
If you would like individual help with writing a data management plan, please contact the Research Data Management Support service at research-data@mdx.ac.uk
For more information about securing computer equipment, please contact the Project Leader: IT Security. (intranet link log in to access)
If you've not written a data management plan before, it can be helpful to look at what a good example should look like.
When using an example plan, consider how the issues raised would apply to your project to ensure that the plans you make are appropriate for the data you will be collecting or creating.