Access and restrictions
How quickly and how broadly to allow others access to your research data are questions that should be answered as part of the data management planning process.
Forms of control
For many researchers, the question is not whether to share data, but how broadly to share: within a closed research group or department, with students, with other researchers in an institution, subject-discipline, country, or with the public at large.
As indicated on the previous page (methods for sharing data), sharing can be done by informal or formal means.
Clearly an advantage of informal sharing is the researcher retains control over who uses the data and how.
With a popular dataset, or over a longer period of time, this may become a disadvantage, and more formal methods involving publishing or depositing the data with a 'trusted repository' or service may be more desirable.
Options for restricting access include:
- Requiring the user to confirm their agreement to follow stated Terms of Use.
- Attaching a legal license to the data, free or fee-based.
- Requiring password protected registration and contact details such as an email address (enforced through automatic email confirmation).
- Requiring authenticated registration for users from certain institutions (for example EASE for University of Edinburgh, UK Access Federation for UK Higher Education institutions).
- Authorising a written statement about how the data will be used for research purposes.
- Publishing a metadata record only (title, author, abstract) with an email contact to request access to the dataset.
- Limiting the number of concurrent users at a given time.
- Limiting access to certain Internet Protocol (IP) addresses.
- Limiting physical access to a secure non-networked server (a data enclave).
Many of the options above will hinder users' ability to discover the existence of the data through search engines.
Open data
Open Data is a philosophy and practice requiring that certain data are freely available to everyone, without restrictions from copyright, patents or other mechanisms of control.
Peter Murray-Rust
Cambridge University (in Wikipedia)
Those who advocate for open data are often interested in new computing techniques for combining, analysing, and integrating online data to create new forms of data or knowledge.
Text mining and "mashups" - such as spatial visualisations enabled through the use of Google maps - are two such techniques which are hindered by restrictive copyright permission, licensing or registration requirements.
The coolest thing to do with your data will be thought of by someone else.
Rufus Pollock
Cambridge University and Open Knowledge Foundation
This thought is exciting to some and disturbing to others. In itself it is a good reason to think through your reasons for restricting your data.
Data protection and confidentiality
If your research involves human subjects, you will need to consider both legal and ethical obligations regarding sharing your data.
The 1998 Data Protection Act affects the processing of personal or sensitive data and the circumstances under which you can share it with others.
The University's Records Management Section has online guidance and offers direct support for decisions about information disclosure.
The role of the University's various research ethics committees is to develop policy and general guidance for Colleges and Schools on ethical issues arising from non-medical research involving human participants. Your School may have its own research ethics committee or guidance.
Avoidance of disclosure of personal or sensitive data can be accomplished in a number of ways, including anonymisation techniques or data aggregation for numeric data, editing of video or sound recordings, use of pseudonyms in qualitative data.
Different methods have different consequences for data quality, and should be considered in tandem with the consent process, for example, what sort of informed consent you seek from your subjects.
The UK Data Archive has an excellent guide on consent, confidentiality and ethics as part of their Managing and Sharing Data guide.
This article was published on Jun 20, 2011