Building inequality data hubs for Africa
7 May 2020
Excellence in data scholarship is essential for research that aims to contribute to evidence-informed policymaking to reduce African inequalities. As a centre of excellence of the African Research Universities Alliance (ARUA), ACEIR has undertaken to help consolidate and broaden the ARUA network by providing ACEIR nodes and their partners with the capacity to build strong national data centres and contribute to ACEIR’s collective analysis and research.
As a technical partner of ACEIR, DataFirst has an extensive history of training and capacity building as well as open data advocacy. The unit has run workshops on data curation throughout the continent and has helped national statistical offices to set up data dissemination portals. It also runs specialist courses on high-level data skills.
Aspiring to the international standards
DataFirst has developed a specific competence in the assessment of data quality issues and in the preparation of diverse data for research use, explains Lynn Woolfrey, the DataFirst manager in charge of data operations.
“DataFirst uses a data life-cycle model which complies with the international standard for Open Archival Information Systems, an ISO standard. This approach is the framework for our data curation services – starting with identifying and sourcing relevant data sets, including liaising with data holders in government and academia.”
Preparing data for reuse involves quality checks and anonymisation of potentially identifiable personal information in the data. A metadata record is prepared for each dataset, including a quality statement.
The datasets are made available publicly (online) for free to researchers except for sensitive, potentially disclosive data, which are available for use at DataFirst’s Centre at the UCT School of Economics.
Setting up ACEIR data hubs
DataFirst will set up a data hub for the South African node of ACEIR, and assist the other ACEIR nodes – based at the University of Ghana, Legon; and University of Nairobi, Kenya – to set up data hubs for their inequality research data output.
DataFirst will encourage an open data approach from ACEIR nodes, explains Lynn.
“The gold standard is open access … accessibility is the cornerstone of data quality and a key principle of DataFirst’s data curation services.”
This principle is in line with the Open Government Partnership, of which South Africa is a founding member. The Partnership assists governments to become “sustainably more transparent, more accountable, and more responsive to their own citizens”. Such transparency and accountability are demonstrated by – amongst others – making public sector data accessible online and by promoting open data practices.
“Open data should not to be confused with open access, which means making the final research publication freely available”, explains Lynn. “Open data practices, on the other hand, mean sharing the data that underlies your research, so that this data is available for researchers’ reuse, scrutiny and feedback.”
Training of ACEIR nodes
The first capacity building training by DataFirst for ACEIR colleagues took place in January this year. Researchers from the nodes, together with Statistics South Africa researchers, received training from DataFirst in using R software for data analysis, with a special focus on spatial inequality mapping and analysis.
The Ghanaian and Kenyan nodes will also receive training in how to prepare their data and create metadata for publishing online, and how to administer the data dissemination software which will be installed by DataFirst for their respective data repositories.
The value of working with DataFirst on the all-important aspect of getting the data on African inequalities right is multi-fold: not only does it give African researchers a wider collection base, but also comes with a higher profile by working with Africa’s only data curation service that is accredited with the CoreTrustSeal certification. Having originated as the African test site for World Bank data repository software in 2009/10, the software that is used enables data quality feedback from researchers – meaning a value add for producers of these data sources.
Article by Charmaine Smith, ACEIR communication manager.