What type of study is this?

September 5, 2025Open Access

Novel network analysis of all research requests in Scotland

STStella Telford RSRosie Seaman JMJennifer Muir

Key Points

The study identified 123 unique datasets requested for research in Scotland since 2015.
Out of these, 114 datasets were not previously available through the Researcher Access Service.
Community detection highlighted groups of datasets, such as education and health datasets, that should be offered together.
The analysis informs how to scale the Researcher Access Service to better meet researcher needs.

Abstract

ObjectivesResearch Data Scotland (RDS) launched the Researcher Access Service (RAS) in 2024. The service was codesigned to meet researcher and data controller needs and maximise the value of and scale administrative linked data in Scotland. Analysis was carried out on all research requests in Scotland since 2015 to identify datasets most commonly asked for in combination to scale the service efficiently. MethodA dataset that included all research requests in Scotland since 2015 was analysed to identify all unique datasets requested for research in Scotland. Network analysis was then used to understand the relationships and connections between datasets. The networks were visualised, and different parameters were used to better understand connections. Community detection was used to identify the datasets which needed to be provided together to maximise the number of projects which could use the Researcher Access Service pathway. ResultsThe work identified that there have been 123 unique datasets requested for research in Scotland since 2015. Of these 123, 114 were not already available through the Researcher Access Service. The top five most requested datasets that were requested in combination with RAS datasets, alongside the number of times requested are as follows: SBR (Scottish Birth Record) – 131 SICSAG (Scottish Intensive Care Society Audit Group) – 84 ECOSS (Electronic Communication of Surveillance in Scotland) – 75 SCI Diabetes (Diabetes Register) – 70 NRS Stillbirths – 64 A number of communities of datasets were identified, for example Education datasets, health datasets, and imaging datasets. Identifying these communities of dataset requests, the need for groups of datasets to be provided together through the RAS became evident. ConclusionThe work has informed decisions on how best to scale the service. Only when we consider the combinations of datasets being requested and the connections between datasets, are we able to design a modern, scalable and sustainable service.

AI에게 질문

Bookmark

View Full Paper