Using Data Science for Evaluation

Speaker(s):

Michele Binci, Kantar Public

Abstract:

This session would be an opportunity to discuss the potential for data science applications in evaluations, with a focus on evaluations in the UK public policy domain. The existence of big data in the UK public sphere (e.g. longitudinal national surveys, administrative datasets, and programme management information systems) offers a unique opportunity to apply machine learning techniques to complement more traditional evaluation methods. In addition, data science applications such as natural language processing, satellite imagery and remote sensing, as well as web scraping can expand the evaluation toolbox at our disposal and enrich our evaluative judgments. We are proposing to discuss all this in the webinar, presenting applications from our current work and our plans for future applications. We would then open the floor to the research and methods expert participating int the conference and webinar to discuss the potential, risks and limitations associated with data science in evaluation.