Datasets

Finding data that you can use in your research can be a challenge. The first step is searching for data but once you find data you will need to evaluate the quality of the data to ensure that it is reliable and suitable for your research. The guidance in this section will help you with step one - searching for data. Guidance on reusing data is also available.

There are many data resources available but a good place to start your search is within your research community and discipline. If a subject specific repository is widely used by your peers, then it may be beneficial to start your search there. If there is a researcher who publishes outputs of interest to you then it my be helpful to look for their datasets in their institutional repository. Most universities will have a research repository or research information system where researchers will have lists of all their datasets. For example, Abertay researchers have their datasets indexed in the Abertay research Portal.

If you are unsure where to start then a simple Google dataset search may be a helpful starting point but just as it is when searching for research papers, no one search will find everything so you will need to carry out many searches and adapt your search depending on the search functionality within each database or search engine.  

Listed below is a list of  some resources that you may find useful when searching for data. 

1. Subject Specific Data Repositories

There are numerous subject specific repositories - too many to list individually. If you are interested in identifying subject specific repositories  have a looks at r3data.org Registry of Research Data Repositories. This will allow you to search or browse for subject-specific repositories that you may want to search within.

 2. Funder Open Access Data Repositories

Many funders do not specify a specific repository for data arising from funded projects. Instead, they require that the repository where the data are to be deposited meets specific requirements. The repositories listed below are useful if you are interested in finding data from specific datasets arising from the funders listed below.

•  The UK Data Service is the UK’s largest collection of social, economic and population data resources. It holds data from studies funded by the Economic and Social Research Council (ESRC)  but also a large number of other Social Science datasets.  If you  plan to search for data in the UK Data Service then you may find the recordings of their webinar training sessions available from YouTube UK Data Service useful.
• Natural Environment Research Council (NERC) funded data can be searched for using the NERC Data catalogue Service. The Five data centres can also be searched or browsed individually. Links to each of the data centres listed below can be found on the Environmental Data Service (EDS).
  • British Oceanographic Data Centre
  • Centre for Environmental Data Analysis
  • Environmental Information Data Centre
  • National Geoscience Data Centre
  • UK Polar Data Centre 
• Zenodo hosted by CERN was launched in 2013 as a result of the EU funded OpenAire project to provide a catch-all repository for EU funded research. it is a good starting point if searching for EU funded data.
• The BBSRC provide a list of data sharing and data resources. 

 3. Useful Dataset Search Engines

No single search engine will find datasets in every repository. Each will be limited to the repositories indexed, and there will be some overlap. This list is not exhaustive.

• Google Dataset Search 
• DataCite Commons - global non-profit organisation that provides persistent identifiers (DOIs) for research data with a goal of helping the research community locate, identify, and cite research data with confidence. DataCite search gathers metadata for each DOI assigned to an object which is used for a large index of research data that can be queried directly to find data
• Data Citation Index –this is part of Web of Science, one of the Library subscribed databases, and access is available from the Library Resources A-Z . A list of the repositories indexed and searchable is available. Web of Science have a useful guide on how to use the Data citation Index and there is also a short You Tube tutorial - Getting Started with the Data Citation Index . 
• OpenAire - a European project supporting the Open Science movement. A comprehensive and open dataset of research information covering 194M publications, 74M research data, 653K research software items. it has a useful browse by United Nations Sustainable Development Goals functionality.
• CESSDA Data Catalogue  Datasets come from over 20 European countries. Good for searching and finding European social science data ( UK data also included). Currently, about 75% of study descriptions are available in English.
• Eu Open Data Portal - search for public data published by the EU institutions, agencies and other bodies.
• EMBL-EBI - part of the European Molecular Biology Laboratory (EMBL), an intergovernmental research organisation funded by over 20 member states, prospect and associate member states.

4. Interdisciplinary Open Access Data Repositories  

• Zenodo -  listed here as an interdisciplinary repository but it also holds data from projects funded by the European Commission and includes UK projects in addition to other non funded datasets
• Figshare can host data deposited by researchers from any institution subject to limits on data size but some institutions also use Figshare as their institutional data repository. 
• Mendeley Data  is a free and secure cloud-based interdisciplinary communal repository. it can also be used to search for datasets deposited in other repositories as it includes a searchable index of open datasets.
• Open Science Framework general data archive
• Dryad repository - All content is licenced CC0 

Data Providers: UK 

• UK Data Service 
• Office for National Statistics
• UK Government Data - Find data published by central government, local authorities and public bodies to help you build products and services.
• Scottish Government statistics
• National Records of Scotland
• The library subject guides will also contain links to some of the most useful resources for your subject. 

Data Providers: International 

• European Statistical Office (Eurostat)
• World Trade Organisation
• World Bank Open Data
• UN Data
• The library subject guides will also contain links to some of the most useful resources for your subject. 

Abertay Datasets

Pure acts as a catalogue for Abertay datasets  and includes datasets deposited in Pure as well as datasets deposited in external data repositories. You can find these listed on Abertay's Pure Portal. Most universities will have an open access repository so if there is a specific researcher or University that you are interested in, check their institutional OA repository for datasets. 

As mentioned previously once you have found data  the next step is to check that it is suitable for your requirements. Reliable data will be accompanied by information that describes the context and provenance of the data. It is important that you ensure that the origin of the data, the creators and methodology are clear and that the rights of the data subjects have been respected. The main challenges to reusing open access data are:

• Data Quality and Integrity: The data may have issues with accuracy, completeness, and consistency. Thorough evaluation of the quality of the data  is essential.
• Metadata and Documentation: Datasets often do not have sufficient metadata and documentation making it difficult to understand the context, methodology and limitations of the data.
• Data Compatibility: Integrating datasets from different sources can be challenging due to differences in formats, units of measurement, and data structures and licensing.
• Ethical and Legal Concerns: There can be issues related to privacy, consent, and intellectual property rights. You must ensure that they comply with all relevant ethical guidelines and legal requirements.
• Version Control: Datasets may be updated over time, and it can be challenging to track changes and ensure that the correct version of the data is being used.
• Trust and Reliability: It is important that the data sources can be verified. Trust in the data's provenance is essential.

Licences: Always check the licence and reuse terms, and provide a citation for the derivative dataset. The licence associated with the original dataset or datasets will determine how the new dataset can be used and shared. If you have used more than one dataset and these have different licences, then usually the resultant dataset must be licenced under the most restrictive licence of the original datasets. More information on licence stacking is provided below.

Commonly Used Dataset Licences

1.  Creative Commons (CC) Licences.

• Creative Commons Attribution licence, CC BY
• Public Domain Dedication, CC0

The CC BY and CC0 licences above are the most commonly used CC dataset licences. They have the least number of restrictions on reuse and are therefore the preferred Creative Commons licences for datasets. More restrictive CC licences include: 

• No derivative Licence  CC BY-ND 
• Non Commercial  CC BY-NC
• Share Alike CC BY-SA 

and combinations of the above licences. 

• CC BY-NC-SA
• CC BY-NC-ND

These more restrictive licences will limit how the data can be reused and combined with other datasets.  Additional information for each licence is available from the Creative Commons website.

2. Other Open Licences

• Open Government Licence
• Open Data Commons Attribution License (ODC-By)
• Open Data Commons Open Database License (ODbL)
• Public Domain Dedication and License (PDDL).
•  Commonly used Software licences and summaries 

The Data Curation Centre have a useful guide  How to license Research Data. if you are reusing datasets with different licences, the OpenMinTed Licence Compatibility Matrix will help you check if the licences are compatible. If you need any advice regarding licences, email repository@abertay.ac.uk.

If you use other people's data to generate a new dataset to you must always acknowledge the original data source just as you would cite any other primary source such as an article, book, image or website. Citing datasets is important  because it:

• Gives credit to the creators of the original datasets
• Promotes data reproducibility
• Allows funders to track the impact and reuse of data arising form their funded studies
• Encourages the citation of data as the normal rather than the exception
• Increases the discoverability.

Usually the elements of a data citation will include:

• Author/Creators
• Year of publication
• Title;
• Publisher (i.e. the data repository)
• Version or edition if applicable
• Permanent identifier for dataset e.g.,  doi

 An example of a dataset deposited in the Dryad repository  is: 

 Cuthill, Innes C. et al. (2017), Data from: Optimizing countershading camouflage, Dryad, Dataset, https://doi.org/10.5061/dryad.rd47f

More information on citing data is available in How to Cite Data Sets and link to Publications available from the Data Curation Centre.