Big Data for SDG3 and SDG5: Promise and Inequality Traps
-
2019•09•10 Kuala Lumpur
Dr Claudia Lopes of UNU-IIGH shared her thoughts on “Big Data for SDG3 and SDG5: Promise and Inequality Traps” as part of the panel of presenters at the Digital Health Seminar on 21 August 2019 hosted by UNU-IIGH.
Dr Claudia Lopes (Photo by UNU-IIGH) Creative Commons BY-NC 2.0
She explained how data sources that are created organically as the by-product of using digital technologies such as the internet, mobile phones and health devices can be leveraged to produce knowledge about the social and physical world in ways not possible using traditional data sources. Adding a dynamic perspective, big data has the potential to uncover longitudinal and large-scale geographical trends as well as co-occurrences and patterns that may reveal aspects of inequality or exclusions of certain populations.
Many lauded on the value of big data as a data source for evaluation, revealing when interventions have worked and predicting their outcomes; however, Dr Lopes cautioned that big data is often biased, incomplete and inaccessible due to commercial interests and ethical and privacy concerns. She pointed out that access to big data sources and technical skills are not equally distributed across regions, and explicated how gendered access and use of technologies imply that big data sources reflect incomplete or distorted realities for men and women. Machine learning models trained on biased data would then reinforce existing inequalities when used as evidence for policy decisions.
She emphasised why we must acknowledge the tension between misuse and missed use of big data, and the importance of citizen generated data to give voice to groups that are not represented in big data sources due digital or social exclusions, to ensure that no one is left behind.
Slides of the seminar: CALopes_Digital Health Week 2019