10th U6+ Consortium of African Universities International Conference

One challenge faced by those who learn and teach statistics in Africa is that most statistics textbooks are written and published in the Global North, and consequently, most of the datasets used in the examples and exercises are from Global North contexts. This creates a challenge in making statistics relevant and interesting to students when the context of the dataset, and other features such as units of measurement, do not speak to the students’ own context. There are reams of interesting and relevant datasets from the African context that are accessible on publicly available webpages, albeit often in unstructured form. These include government and private sector websites. Web scraping, using readily available functions in programming languages such as R and Python, allows web users to extract data from webpages and wrangle them into structured datasets. By doing so, lecturers and teachers of statistics can curate datasets that are contextually relevant to their students and are recent, thus making statistics education in Africa more interesting and inclusive. This talk will present a few case studies of interesting datasets curated using web scraping, from domains such as climate, retail, and sport, and will provide an overview of some of the technical and ethical issues that would-be web scrapers should be aware of.