In the first café conversation of 2021 we talked about some of the RDM practices followed by our researchers as well as some of the challenges they face. RDM is a complex subject with many discipline-specific issues and attitudes. Some researchers produce voluminous amounts of data (many Terabytes) while others are more concerned about ethics, data protection and data anonymisation. Publishing research data is expected in some disciplines while in others it may be considered a low priority without any incentives. However, the overall trend is towards Open Scholarship, which leads to more robust and transparent research because anyone (other researchers, the commercial sector, the public) can scrutinise research data and attempt to verify findings. Open Scholarship can lead to making more progress as a research community collectively in an attempt to solve difficult research problems and to learn more about the universe.
We spoke briefly about the meaning of “research data”. One broad definition, according to the Digital Curation Centre in Edinburgh, is anything that can be used to validate or replicate a research conclusion, or enrich understanding of the research process. In addition to this, most digital research data exists in three forms at the very least: raw data, processed data and published data. All research data needs to be managed responsibly because this reduces the chances of inadvertent data leaks or loss.
Practically all researchers need a storage solution for their research data and we discussed this at some length. There are many short-term and long-term storage solutions available, each with strengths and weaknesses.
As for short-term storage, the University’s Computing and Information Service (CIS) provides OneDrive for Business, Sharepoint, PRS (Personal Research Storage) and SRS (Shared Research Storage). There was general agreement that OneDrive and Sharepoint are easy to use, particularly if you need to share files with external collaborators. PRS and SRS are good solutions but more difficult to set up and use. Each storage solution is designed with different requirements in mind. [CIS have developed a Storage Options Tool to help researchers choose the best solution depending on their requirements.]
Departmental servers and storage were mentioned several times during the conversation. Generally, these are managed by CIS except for ICC (Institute for Computational Cosmology) which has its own managed infrastructure for historical and technical reasons. We heard during the conversation that departmental servers and storage are used for both short-term and long-term storage. There was some concern that not all research data were being backed up regularly, possibly due to the volume of data.
As for long-term storage, we spoke a little about the University’s Research Data Repository. Not everyone attending the café conversation knew the University had its own data repository. It was developed prior to May 2015 to support a new EPSRC research data policy which began at that time. The University has tentative plans to combine its various repositories together into one multi-purpose repository and to integrate it with other University systems (e.g., Worktribe). We also spoke briefly about the OSF (Open Science Framework) repository, the Zenodo repository and the UKRI repositories. OSF and Zenodo are both multi-disciplinary repositories whereas the UKRI repositories are subject-specific repositories (e.g., ESRC’s ReShare repository and NERC’s UK Polar Data Centre). At least one researcher attending the conversation asked for more University guidance on choosing a repository for their research data.
One recurring topic throughout was on University guidance and RDM support, particularly for graduate students and new staff. Where to seek guidance and support? What to do with student research data? What to do with paper documents containing research data? How to hand over research data when a researcher leaves the University? We can mount OneDrive on the Hamilton supercomputer, could CIS enable us to mount OneDrive on SRS? These were some of the questions raised during the conversation.
We spoke about storing the process of research. Nvivo and OneNote were mentioned as tools for recording research ideas and processes. This topic, I believe, is just one aspect of the much larger topic, best practices for scientific computing. We have two, upcoming café conversations in which we can discuss aspects of this broad topic at greater length:
27 January: Quality control and Software Development Practices
03 February: Sharing Best Data and Project Management Practices for Organising Data and Avoiding Digital Clutter
Please come and join us, if you can.
Nicholas Syrotiuk (@DurhamRDM) is the University’s Research Data Manager.
------------------------------------
This blog expresses the author's views and interpretation of comments made during the conversation. Any errors or omissions are the author's.
We spoke briefly about the meaning of “research data”. One broad definition, according to the Digital Curation Centre in Edinburgh, is anything that can be used to validate or replicate a research conclusion, or enrich understanding of the research process. In addition to this, most digital research data exists in three forms at the very least: raw data, processed data and published data. All research data needs to be managed responsibly because this reduces the chances of inadvertent data leaks or loss.
Practically all researchers need a storage solution for their research data and we discussed this at some length. There are many short-term and long-term storage solutions available, each with strengths and weaknesses.
As for short-term storage, the University’s Computing and Information Service (CIS) provides OneDrive for Business, Sharepoint, PRS (Personal Research Storage) and SRS (Shared Research Storage). There was general agreement that OneDrive and Sharepoint are easy to use, particularly if you need to share files with external collaborators. PRS and SRS are good solutions but more difficult to set up and use. Each storage solution is designed with different requirements in mind. [CIS have developed a Storage Options Tool to help researchers choose the best solution depending on their requirements.]
Departmental servers and storage were mentioned several times during the conversation. Generally, these are managed by CIS except for ICC (Institute for Computational Cosmology) which has its own managed infrastructure for historical and technical reasons. We heard during the conversation that departmental servers and storage are used for both short-term and long-term storage. There was some concern that not all research data were being backed up regularly, possibly due to the volume of data.
As for long-term storage, we spoke a little about the University’s Research Data Repository. Not everyone attending the café conversation knew the University had its own data repository. It was developed prior to May 2015 to support a new EPSRC research data policy which began at that time. The University has tentative plans to combine its various repositories together into one multi-purpose repository and to integrate it with other University systems (e.g., Worktribe). We also spoke briefly about the OSF (Open Science Framework) repository, the Zenodo repository and the UKRI repositories. OSF and Zenodo are both multi-disciplinary repositories whereas the UKRI repositories are subject-specific repositories (e.g., ESRC’s ReShare repository and NERC’s UK Polar Data Centre). At least one researcher attending the conversation asked for more University guidance on choosing a repository for their research data.
One recurring topic throughout was on University guidance and RDM support, particularly for graduate students and new staff. Where to seek guidance and support? What to do with student research data? What to do with paper documents containing research data? How to hand over research data when a researcher leaves the University? We can mount OneDrive on the Hamilton supercomputer, could CIS enable us to mount OneDrive on SRS? These were some of the questions raised during the conversation.
We spoke about storing the process of research. Nvivo and OneNote were mentioned as tools for recording research ideas and processes. This topic, I believe, is just one aspect of the much larger topic, best practices for scientific computing. We have two, upcoming café conversations in which we can discuss aspects of this broad topic at greater length:
27 January: Quality control and Software Development Practices
03 February: Sharing Best Data and Project Management Practices for Organising Data and Avoiding Digital Clutter
Please come and join us, if you can.
Nicholas Syrotiuk (@DurhamRDM) is the University’s Research Data Manager.
------------------------------------
This blog expresses the author's views and interpretation of comments made during the conversation. Any errors or omissions are the author's.
Very useful information. Thanks for sharing
ReplyDeletehttps://aditidigitalsolutions.com/data-science-training-hyderabad/