Along with these new types of data big data I started to wonder about sources of what I used to think of as big data – national surveys and longitudinal studies and administrative data collected by Government. What Will Parry refers to as big(ish) data (in this handy post on relatively simple approaches for reducing the size and complexity of data).
Administrative data was defined by the UK Administrative Data Liaison service as:
“…information collected primarily for administrative (not research) purposes. This type of data is collected by government departments and other organisations for the purposes of registration, transaction and record keeping, usually during the delivery of a service.”
As with big data, there are many advantages of using administrative data for research, including the cost savings, regular updates, being able to link different datasets and capture populations who may not respond to a survey or other methods and identifying counterfactual and control groups after an intervention. And there are similar challenges, around access, data protection, missing data and other issues when data is not designed for the purpose for which it is used.
But how do you get hold of it?
The Administrative Data Research Network is a UK-wide partnership between universities, government departments and agencies, national statistics authorities, the third sector, funders and researchers. Researchers have to be accredited to access linked, de-identified administrative data. The Administrative Data Service coordinates the Network and is the first point of contact for researchers who want to use administrative data.
Other data is more open and immediately accessible through the UK Data Service (although some sources still require registration to access). Their collection includes major UK government-sponsored surveys, cross-national surveys, longitudinal studies, UK census data, international aggregate, business data, and qualitative data. The site also provides advice and training. These data sources can be useful for a wide range of socio- economic research and evaluation.
Government is increasingly trying to make data available, for research and other purposes, and this site, although still under development, and this blog are ones to watch.
New Philanthropy Capital’s Data Labs can help some not-for-profit organisations to access government administrative data and better understand the impact of their services. While this is currently limited to the criminal justice sector, it is likely to expand rapidly for education and health organisations.
As Adam Steventon, Director of Data Analytics at the Health Foundation, notes here, large administrative data sets have been used since the 1950s, the progress and potential being in better data linkage and analysis. Efforts are being made to support increase value from data through better linkage. CLOSER, in addition to improving access to the UKs longitudinal studies, has a work stream focused on linking the studies to other records. For more on data linkage listen to this podcast from the National Centre for Research Methods.
Better access, linkage and analytic possibilities for administrative data is something to think about if you are commissioning or delivering analysis or evaluation for an organisation. What access and knowledge do you or your provider have about these sources of administrative data? What skills to get the best from it? And, if commissioning, how comfortable are you interpreting what you are given, or interrogating how others have interpreted data? For addressing many of these questions this course (or something similar), which I cover in detail in my next post, would be a good place to start.