To cite this article:
Mine Dogucu & Mine Çetinkaya-Rundel (2021) Web Scraping in the Statistics and Data Science Curriculum: Challenges and Opportunities, Journal of Statistics and Data Science Education, 29:sup1, S112-S122, DOI: 10.1080/10691898.2020.1787116
This paper is now published online on Journal of Statistics Education.
Best practices in statistics and data science courses include the use of real and relevant data as well as teaching the entire data science cycle starting with importing data. A rich source of real and current data is the web, where data are often presented and stored in a structure that needs some wrangling and transforming before they can be ready for analysis. The web is a resource students naturally turn to for finding data for data analysis projects, but without formal instruction on how to get that data into a structured format, they often resort to copy-pasting or manual entry into a spreadsheet, which are both time consuming and error-prone. Teaching web scraping provides an opportunity to bring such data into the curriculum in an effective and efficient way. In this paper we explain how web scraping works and how it can be implemented in a pedagogically sound and technically executable way at various levels of statistics and data science curricula. We provide classroom activities where we connect this modern computing technique with traditional statistical topics. Lastly, we share the opportunities web scraping brings to the classrooms as well as the challenges the instructors and tips for avoiding them.