The Network has experts across all aspects of using data for research, including data de-identification, data linkage, legal issues for the ADRN researcher and public engagement. They have collaborated to create guides to help researchers use data more effectively.
This guide is designed to give readers a practical introduction to data linkage and is aimed at researchers who would like to gain an understanding of data linkage techniques, either for the creation or analysis of linked data. It covers data preparation, deterministic and probabilistic linkage methods, and analysis of linked data, with examples relevant to health and other administrative data sources. This guide is relevant for academic researchers in the social and health sciences or those who work for government, survey agencies, official statistics, charities or the private sector.
Readers should use this guide as an introductory text on data linkage; to gain knowledge of the background and theory of data linkage methods; to understand how to perform basic deterministic and probabilistic linkage; and to consider how to evaluate and report data linkage quality.
This guide requires no previous knowledge or experience of the topic. It is aimed at academic researchers who have an association with the Administrative Data Research Network as well as at a general audience interested in the subject matter. For Network researchers, the document will serve as useful background for the legal aspects of the certification training they receive before they can access the service. Wider audiences might be interested in how the Network’s secure environment ensures the data access we allow is fair and lawful.
The guide sets out the legal background to data protection laws in the UK, and offers a broad explanation of the current law relating to data sharing and linkage, as well as a consideration of the implications of the impending EU General Data Protection Regulation 2016. There is also consideration of some non-legal issues surrounding the topic. Readers can refer to the Network’s website for further information, and should consult professional legal advice on any specific legal points.
This guide is intended for researchers who need to learn what Output Statistical Disclosure Control is and how that relates to their work, data owners who are concerned about inadvertent disclosure of results and the wider public who are interested in how the ADRN’s security model (the ‘Five Safes’) works to encourage research while protecting confidentiality
This is not a comprehensive guide to Output Statistical Disclosure Control: researchers who use the ADRN and therefore have access to the source data will go through an extensive training programme. The aim of this guide is to provide an overview.
This guide is relevant for academic researchers accessing the ADRN and those in the wider social and economic sciences community. It may also be of interest to those who work in government, survey agencies, official statistics, charities or the private sector and are interested in learning about data quality issues affecting administrative datasets.
The varied nature of administrative data sources and the varied analytical applications of these data mean that the options for dealing with the data quality issues may be both dataset-specific and user-specific. It is therefore not possible to give a definitive list of data quality issues or a definitive list of remedial actions in this guide because each case needs to be considered individually. So this guide aims to raise users’ awareness of the types of potential data quality issues affecting administrative data; help users choose the most appropriate form of administrative dataset for their needs and give examples of ways in which users can deal with data quality issues at various stages of the analytical process.