It is good practice to have a README file to accompany your dataset. A README file should be a txt file.
We would recommend the following guides to writing README files:
To deposit your data and request a DOI that you can cite in your thesis and future publications:
On receipt of confirmation that the data is final and can be released, the research data record will be checked, validated and the DOI will be registered the next working day.
The title should reflect the content of the dataset. It can reference the title of the thesis, but should not be identical.
This should provide more information on the content of the dataset but should include a reference to the Thesis.
This can be a specific date or a range of dates to reflect the period of time during which you collected and analysed the data.
Add anyone who helped in the creation of the data or contributed and, generally your supervisor. Change the default ‘role’ under the edit option once the name has been added. Only Creators will be in the citation on ePrints Soton.
This will reflect the organisational unit that you are attached to on the University system and is filled in automatically.
University of Southampton is the default publisher and should not be altered.
Leave this field blank. This field will be completed by the Research Data team when the DOI is to be registered. A temporary note giving the assigned DOI will be added to the description, then remove prior to the DOI being registered.
Upload files by browsing or dragging and dropping. Add any agreed embargo to the files. Do not change visibility of the files.
The year should be the year that data will be released openly or available on request. If all files are embargoed the year will be the year the embargo lifts.
Only complete this field if the data is about a specific period of time. For example if the data related to the London 2012 Olympics you would add the relevant dates this took place (27 July to 12 August 2012).
This should only be completed if the data is about a specific location. For example if the data is about the Mekong Delta you could add this in words, use of polygon coordinates, point coordinates (10.033333, 105.783333).
If the data collected is subject to any legal or ethical constraints this information should be added.
It is helpful to add a few words that describe the subject of the data
The default ‘Public – no restriction’ should be accepted for data that is going to be
For all other visibility settings contact firstname.lastname@example.org to discuss before adding files to the record.
The fields marked with an asterisk (red asterisk in Pure) are mandatory fields and must be completed before the record can be saved. Once the DOI has been registered they cannot be changed.
The research data that supports your thesis is an essential part of the work that you have carried out. Research data can be an output of research that can bring its own rewards via citation, and open doors to future collaboration. By depositing your data you are able to properly cite the data in publications, including your thesis, using a Digital Object Identifier (DOI). Publishers now often require that research data is deposited in support of publications and funders view it as good research practice to include a data access statement on papers stating how the data can be obtained - see https://library.soton.ac.uk/researchdata/access-statements for further guidance.
The University of Southampton Research Data Management Policy, which applies to postgraduate research students, requires that research data underpinning publication should be deposited in the Institutional Repository (via Pure) and this includes Doctoral theses.
Research data, defined as "material intended for analysis", that directly underpins the assertions, figures and diagrams in your thesis should normally be deposited as a dataset, accompanied by any necessary documentation. This applies to data or research materials that you have collected or created yourself.
It may not mean that every piece of data that you have collected should be deposited. It is important to evaluate the data, something you should do throughout the research process, and remove data from the final dataset that covers aspects of your research that no longer form part of your thesis. It may not be a case of discarding that data, just removing it from the dataset supporting your thesis. See https://library.soton.ac.uk/researchdata/selection for further guidance.
If you have extracted data from commercial databases; obtained data under licence; collected data from archives for personal research purposes only, are using confidential data (for example, MRC or NHS data) or are using material still under copyright, you may not be able to deposit the data. Contact email@example.com for guidance.
Data should be anonymised or psuedonymised before deposit, unless you sought permission from your participants to include their names.
We strongly encourage research data to be as "open as possible, and closed as necessary". However, not all data can be made open. If your data is subject to a commercial confidentiality clause, is sensitive or there is too high a risk that the data can be re-identified then the data does not need to be open. It is possible to make the data available:
Further guidance is available from https://library.soton.ac.uk/researchdata/sharingtips and https://library.soton.ac.uk/researchdata/restrictingaccess. Contact firstname.lastname@example.org for advice if required.
Our Institutional Repository can only accept files or zips up to 4GB in size. We recommend that no more than 10 such files are uploaded to a single record. If your data is much larger than this you contact email@example.com for advice.
If you have more than 10 files we would recommend that the files are placed in a zip folder. It can be useful to zip files of the same type together. For example if you have audio files and analysed data files, these could be in two separate zip files as this could make them easier to describe. However, the data should be zipped in a way that helps to keep the data together in a logical way. In all cases the README should explain the content, the relationship between files and any software required.
Further guidance is available on the main Research Data Management Pages -